Lithiumion technology has already taken the portable battery world by storm.
Figure 1. The Rocking Chair concept
Now research engineers are working to bring this technology to the world of large stationary batteries – the type used for standby power and large-scale energy storage systems.
Lithium is in many respects an ideal material for use in batteries. It is not only one of the lightest elements, it is also highly reducing, giving a high electrode potential that allows cells with operating voltages up to 4 V or more to be produced, as opposed to 1.2-2.0 V for most commercially available cell types. Low weight and high voltage combine to give batteries with very high energy density.
Unlike other battery technologies, such as lead-acid, nickel-cadmium and nickel-metal hydride, lithium-based battery chemistry is non-aqueous (indeed, water is not stable at the high operating voltages of these cells). This eliminates losses due to electrolysis and allows a faradic charge efficiency that is very close to 100 per cent.
Early rechargeable lithium batteries incorporated metallic lithium negative active material. Such electrodes had poor cycling characteristics due to the formation of dendrites. Worse, highly reactive powdered lithium could be formed, sometimes leading to battery fires. One solution to this problem was found as recently as 1990, when Sony first published results of its work based on the insertion of lithium ions into carbon negative electrodes. The insertion of lithium ions into a metal oxide lattice was already the basis of the operation of the positive electrode, so this marked the commercial birth of the “rocking chair” concept. This simply involves the exchange of lithium ions between the electrodes as the battery is charged and discharged.
For the most part, positive active materials used in portable batteries are lithiated cobalt oxides (LiCoO2). Cobalt oxides can be fully lithiated (an important feature for cell manufacturing and operability), have a high specific capacity and a high operating voltage. They give good cycling performance and are reasonably stable at higher temperatures. The main drawback is the high cost of cobalt.
Negative active materials are various types of carbon, either crystallized or amorphous. Graphite has been chosen by a number of manufacturers for its excellent characteristics: low cost, low voltage (resulting in high cell voltage), high specific capacity, and high reversibility.
Electrolytes are mixtures of lithium salts, typically LiPF6, with organic carbonates. Each manufacturer must match the electrolyte formulation to the materials used in the electrodes. Of particular importance in this relationship is the stability of the passivation layer formed on the surface of the negative. If this layer is not stable, the electrolyte and the carbon material will react with one another, thus shortening the life of the cell. One drawback of non-aqueous electrolytes is their relatively low conductivity, about one tenth as high as that of aqueous solutions. This is overcome by the use of thin electrodes.
Historically, most portable lithium ion cells have been cylindrical, with wound electrode stacks. More recently, however, there have been increasing numbers of prismatic cells produced, using flattened electrode coils.
Portable lithium ion cells generally have a maximum capacity of about 5 Wh. In scaling them up for stationary applications, three main challenges must be addressed: the cost and characteristics of the positive active material; construction issues; and safety concerns.
Cobalt is expensive. This is not a major concern for small cells, where assembly costs are often as important, if not more so, than raw material costs. As larger capacity cells are produced, material costs become more prominent and cobalt becomes too expensive to be cost effective. Much research in the industry has therefore focused on the substitution of cobalt by other transition metal oxides, such as manganese, nickel or vanadium. Saft is currently conducting work on doped oxides, in which a portion of one metal is substituted by another.
Building a higher capacity battery is not simply a case of scaling up the existing components of smaller batteries. For example, although prismatic lithium ion cells now represent 50 per cent of the portable market, this construction is less suitable for larger cells. Also, while these cells are not pressurized by internal gas, it is nevertheless desirable to have a reasonably strong container that can compress the plate stacks evenly. Thus, the cylindrical format is more strongly favoured in larger cells.
Large lithium ion cells are generally assembled into modules comprising several cells and their associated electronics.
Initially, the main safety concern with rechargeable lithium batteries was associated with metallic lithium negatives. This has been overcome in lithium ion technology by the adoption of carbon-based negative materials. The remaining concern relates to overcharging. If these batteries are overcharged at high rates and the temperature exceeds about 150°C, a fire will result. The organic electrolyte, negative active material and separators are combustible, so this is undesirable, to say the least.
This problem is avoided by the use of electronic controls during charging. These controls are also required to prevent deterioration in the active materials that would result from overcharging. At least two levels of controls are employed:
- The first level of control operates on each cell or grouping of parallel cells. It either switches the cell out of circuit or shunts excess charging current once the cell reaches the maximum charge voltage. This ensures that all cells reach the same state of charge without overcharging.
- The second level limits the overall charge current to a maximum level which prevents the temperature rising to a dangerous level.
These redundant levels of control make the operation of lithium ion batteries as safe as, if not safer than, any other type. Additional protective devices include fusing and safety valves. Apart from the end-of-charge control deployed at the cell level (which is more an integral part of normal operation than a safety feature), it should be noted that all of these safety devices are to prevent damage caused by external means, such as a charger failure.
Energy density: This is an area in which lithium ion batteries excel. The ability to pack the maximum capacity into the smallest possible space is of vital importance, not just in portable uses like cell phones, where battery capacities are measured in milliwatt hours, but also in energy storage applications, where multi-megawatt hour capacities are expected to be quite normal.
Figure 2. Discharge characteristics of 44 Ah high energy lithium ion cell
Discharge characteristics: lithium ion batteries are capable of providing excellent capacity availability at high discharge rates. Figure 2 shows the discharge characteristics of a 44 Ah high energy cell type and demonstrates that over 90 per cent of rated capacity is available at discharge rates approaching 1.0 C5.
Figure 3. Saft lithium ion batteries
Cycling: Energy storage applications such as load levelling have an obvious need for good cycling capability. lithium ion provides very good deep discharge cycling, with a capability of around 3000 cycles of 80 per cent depth of discharge (DoD). Furthermore, developments in active materials are bringing continual improvements in this area. In shallow cycling, the performance of this technology is outstanding. Figure 4 shows results of cycling smaller (700 mAh) Saft cells at ten per cent DoD and 40°C. The cells are given a 100 per cent diagnostic discharge every 10 000 cycles. After an initial increase in impedance, the cell characteristics have stabilized and there has been almost no capacity fading.
High and low temperature operation: lithium ion batteries can function over a wide temperature range with current designs capable of operation down to around -20°C. Future work should extend this range to even lower temperatures.
At the high end of the range, life testing has shown remarkably little dependence on temperature with very good capacity stability demonstrated when cycling at temperatures as high as 60°C. This is obviously a strong point for high-temperature applications. If a battery room had to be air-conditioned to produce the desired operating life, it would increase the life cycle cost substantially.
Float charging and ‘smart’ batteries: In electric vehicle operation, for which lithium ion batteries are also being developed, there is extensive communication between the battery and charger, through a sophisticated command module. One of the tasks to be faced in developing lithium ion batteries for stationary applications in general is that they must be made compatible with existing “dumb” constant potential chargers, where no communication is possible between the two. For such systems, the electronic controls must be modified to make the batteries Osmarter?. Since the individual cell voltages are being checked by these controls, it is just a small step to turn this into an on-board diagnostic and monitoring capability, allowing remote operator access and minimizing maintenance costs. Lithiumion batteries can be charged quickly and at 100 per cent faradic efficiency using constant potential chargers. Initial charge currents up to 0.2 C5 can be used. Once a cell is charged, the float current is extremely low, less than C/10000. There is no heat evolution, and thermal runaway cannot occur at constant voltage.
Parallel operation: One of the features of lithium ion cells is that their open circuit voltage is directly related to the state of charge (SOC). This provides a very useful diagnostic regarding the state of health of a cell. It also ensures that cells charged to the same end-of-charge voltage will be at the same SOC.
This voltage behaviour enables individual cells to be connected in parallel within a string. Since all parallel cells are charged to the same end-of-charge voltage, it is known that they are all at the same SOC. The same cannot be said of aqueous battery technologies, in which the parallel connection of cells within a string is actively discouraged. In the 6-cell module shown in Figure 3 it is possible to connect the cells as follows: 6 in series; 3 in series, 2 in parallel; 2 in series, 3 in parallel; or all 6 in parallel. Electronics to control the end-of-charge voltage are implemented for each parallel grouping of cells (or on each cell if all are in series).
This arrangement allows considerable flexibility in configuration, while minimizing the number of cell types that must be produced. In turn, this allows more standardization in production and has a direct impact on the cost of this technology .
Service life: The driving force behind the development of large lithium ion cells has been the electric vehicle market. Thus, it is understandable that most of the work on characterizing the life of this product has related to operation under cycling conditions. The results of this work are directly applicable to load levelling applications, where batteries are deep cycled on a daily basis. However, they do not reflect the realities of battery operation in spinning reserve or peak shaving applications, in which much more time is spent on float.
Based on float testing results with earlier electrode materials and extrapolating the initial behaviour of the latest Saft product, we estimate that a life of at least 15 years should be possible under float charging conditions. As already detailed, this is not expected to vary significantly with operating temperature.
Availability and cost issues
The larger lithium ion cell types are already being produced on a pilot plant basis, but it is expected to be another three or four years before industrial production will begin.
In general, none of the raw materials has an intrinsically high cost. However, many of them are unique to this technology and have no other industrial application. The result is that the cost of these batteries is quite sensitive to volume. While portable lithium ion batteries remain expensive compared to other technologies, the production cost has fallen about 30 per cent in the last two years. This is expected to continue, particularly as production picks up for larger cells.
Eventually, cost studies predict that lithium ion batteries will have an equivalent life cycle cost to lead-acid types. However, this is likely to take several years in addition to the three or four years required for the start up of industrial production. We predict that the early uses of this product will be for trial installations; for applications where space is severely limited; or for operation under demanding conditions, where existing battery options cannot provide adequate service.
Energy storage is a broad term that covers a variety of applications and battery needs. The following summarises some of the characteristics of lithium ion batteries, and how they will relate to various application segments.
- High energy density: For all installations, but particularly those with multi-megawatt hour capacities, lithium ion will occupy just one quarter the space of VRLA batteries of the same capacity.
- High performance: High energy lithium ion batteries provide excellent performance for spinning reserve operation. High power versions will be available for power quality applications requiring short bursts of energy.
- Excellent cycling capability: This is desirable for all energy storage applications, but particularly for load levelling and frequency support.
- Zero routine maintenance: Being hermetically sealed, lithium ion batteries require no replenishment of electrolyte. Built-in monitoring and diagnostic capability eliminates routine surveillance. This is a plus for all stationary applications.
- High charge efficiency: For load levelling, the 100 per cent faradic charge efficiency of lithium ion means that energy losses are minimized.
- Excellent operation at temperature extremes: The need for heating or cooling is minimized, thus minimising operating costs.
With general availability a few years away, and with high costs initially, lithium ion batteries for stationary applications still have a long way to go. However, their particular advantages for energy storage applications are likely to make them a strong contender for this market.
This article is based on a paper presented at the 6th International Conference on Batteries for Utility Energy Storage, held in Gelsenkirchen, Germany on September 21-23, 1999.