Technology

(only broad overview and intro no single item working principle) e.g. Sensor fusion issues etc.

Once in a while in daily life situations probably everyone has come across some sort of battery. Everyday battery knowledge divides all batteries into two groups: one time use and rechargeable batteries. In engineering, one time use batteries are in primary battery group while secondary batteries are rechargeable ones. Electric vehicles (EV) use secondary batteries (except for small toy vehicles and special applications) hence in this chapter the term battery means secondary battery unless noted otherwise. From economic perspective, batteries are a serious business approaching 100-billion-euro market size. About a third of all batteries are used as automotive traction batteries (EV, HEV), other third is used in industrial applications and portable applications (consumer electronics) while the last third is used in other applications like power tools and conventional car batteries.

From everyday knowledge it is know that batteries have different voltages. A wall clock typically uses an AA or AAA size 1.5V battery while a car has 12V lead-acid battery under the hood. There are two reasons for different battery voltages: chemistry and series connection. The chemical composition of battery materials determines the voltage in the range of 1.2V to 3.9V. How come a car lead-acid battery has 12V? It actually has multiple smaller batteries inside and they are series connected (mind the polarity) to sum up their voltages. These individual internal batteries are called cells. Figure 1 shows some multi-cell batteries. It would be technically correct to say that a battery is in fact two or more series connected cells of the same kind. Hence a battery composed of just a single cell would not be a battery but rather just a cell. However, to not cause confusion it is accustomed in everyday language to use the term battery for any number of cells while a cell means a single element. This notation will be used here as well. One of-the-shelf battery is the car lead-acid battery which has six 2.1V cells inside (the voltage is rounded to 12V for convenience), another multi-cell battery example is the 9-volt battery which is composed of six 1.5V cells (alkaline or carbon-zinc chemistry). When one installs two AA batteries in a TV remote, they are series connected to form a 3V battery.

In electrical engineering a battery is recognized as a voltage source. A major difference is that the voltage of this source will gradually decrease when a load is applied (discharge) while connecting a battery to a higher voltage source will cause its voltage to gradually increase (charge up). A more precise definition claims that a battery is in fact an electrochemical device which can provide voltage and release electrical energy stored inside of it in the form of chemical bonds.

This electrochemical device has four main parts: two electrodes, electrolyte and separator. Both electrodes typically are solid materials made to store charge and conduct current. One of the electrodes is cathode while other is anode. From the basics of electronics, current flows out of the cathode and into the anode. This holds true for the battery as well – however only if the battery is being discharged. When the battery is being charged, the current flows in other direction, hence cathode and anode switches places. This can cause some confusion. To avoid it, in electronics battery electrodes are typically denoted by their polarity: electrode with higher polarity is the positive electrode (positive terminal, positive pole or just +) and electrode with lower polarity is the negative electrode (negative terminal, negative pole or just -). The switching of anode and cathode can cause confusion even in the field of battery chemistry. Probably due to the fact that batteries are most useful when they are being discharged (and on stand-by) battery chemists have fixed anode and cathode to this position: anode is negative and cathode is positive. The next essential part of the battery is the electrolyte which typically is a liquid although a major research is performed to develop and introduce solid-state materials. Electrolyte connects both electrodes to provide ion transfer between them during charging and discharging while inhibiting internal flow of electrical current. The electrical current, in the form of electrons, flows through an external circuit. If both electrodes inside the battery would come into direct contact, a short circuit would form which would lead to a failure of the battery. A separator layer is introduced to provide electrical and mechanical isolation between the electrodes. The separator typically is a solid state material with micro pores to allow ion transfer. Some of modern batteries are made of three distinct layers: anode, electrolyte soaked separator and cathode. Other parts of the battery include the case with markings, external isolation materials like plastic coating, current collectors and external terminals or connectors.

Batteries have a wide array of different technical parameters: electrical, physical, chemical and sustainability. The most critical parameter is the battery chemistry. It defines all other parameters of the battery. Chemistry in regard to batteries describes composition of both electrodes and the electrolyte. Some examples are lead-acid battery, nickel-metal hydride (NiMH) battery and many types of lithium-ion (Li-ion) batteries which can be further classified by cathode and anode materials.

The chemical composition of electrodes defines the voltage of a single cell. All types of battery cells have a certain nominal voltage U_nom. As previously noted, the nominal voltage of different chemistries is in the range of 1.2 V to 3.9V. The nominal voltage is somewhere between maximal voltage U_max (charging voltage) and minimal voltage U_min (discharge cut-off voltage, end-of-discharge) and it is defined by the manufacturer of the cell. The nominal voltage is used for calculations to determine the voltage of the battery pack if cells are series connected. It is also used to calculate the watt-hour (Wh) capacity of cell or whole battery pack. Nominal voltage of a single lead-acid cell is 2.1V, of NiMH cell it is 1.2V and for majority of Li-ion cells it is 3.6V. Charging voltage is the maximal voltage which can used to charge the cell. Higher voltage leads to overcharging and often to harmful effects which degrade the battery. Discharge cut-off voltage is the voltage beyond which discharge should be terminated to prevent damage to the cell. At this point it should be clear that battery voltage is not a fixed value – it changes depending on the depth-of-discharge (DoD). The fuller the cell/battery the higher the voltage and vice versa. A battery discharge voltage curve is given in figure 2. It can be seen that for most part the voltage is changing little as the cell is being discharged. This is regarded as the flatness of the voltage curve. For primary batteries it is desirable to have a flat curve which translates to stable supply voltage.

The second most important quantitative battery parameter is capacity Q_bat. Capacity determines how much charge a battery can store. It is measured in amp hours (Ah). Higher Ah rating means the battery will be able to run longer before requiring a recharge. If the load current I_load is known then the runtime t can be calculated as follows:

For example, a battery rated at 3Ah will be able to deliver 1A for three hours or run for one hour while delivering current of 3A. It can be seen that change in current results in inversely proportional change in discharge time. If one wants to compare two batteries with identical nominal voltages then Ah capacity rating can be used to determine which battery can store more charge. For example when selecting a lead-acid battery for car, the voltage of the battery is determined by the car: it has to be 12V battery. A higher capacity battery can be used to improve performance. However if an engineer has to select a battery for an application which can use different nominal voltages then energy rating E_bat (measured in watt hours Wh) should be used instead of capacity. While nominal voltage and capacity is typically printed on the battery label, the energy rating is only occasionally given on some Li-ion cells. The energy rating can be obtained by multiplying capacity and nominal voltage:

For example, a 10Ah lithium-ion cell with 3.6V nominal voltage will be able to store 36Wh of energy. A 3Ah lead-acid battery with 12V nominal voltage will also store 36Wh which means that both batteries have the same energy rating albeit at different nominal voltages. Both capacity Q_bat and energy E_bat ratings have time (expressed in hours h) in their definitions. This can be used to calculate battery runtime if the load power P_load is known:

For example, if a battery has 200Wh of energy and it is connected to a constant 50W load then the battery will be fully discharged after 4h. If an engineer needs to select a battery to supply 100W load for at least a full day (24h) then she should select at least 2400Wh (2.4kWh) battery. However, for these equations one has to remember that rarely the load is constant power or constant current. Moreover, as the battery discharges, its voltage will decrease which will lead to changes in load current and power in most cases.

The next electrical parameter is current. A good battery datasheet will provide at least a few current values at different conditions. Common parameters are standard charge current, rapid charge current, max. continuous discharge current and standard discharge current. Often the charging current ratings are significantly lower than discharge ratings. Battery lifetime will be improved if it is used at standard ratings. Standard ratings are often used to determine nominal capacity – at increased currents the battery will not be able to deliver its nominal capacity. The rapid charge current and max. continuous current typically depends on the thermal performance of the battery cells, especially in the case of Li-ion. A high rate discharge/charging will lead to heating of the cell – once temperature gets too high the current has to be stopped to prevent damage to the cell.

In engineering and battery datasheets there is another battery-specific parameter which is directly related to Ah rating: the C-rate. The value of 1C is a number same as the nominal capacity of the battery. The C-rate itself has no unit of measurement but when it is converted to current it is expressed in amps A. C-rate is used to determine current for both charge and discharge. It comes handy when comparing current capabilities of different batteries and simply estimating how large the current is in respect to capacity of the battery. For example 2C discharge rate of a 10Ah battery is 20A while 0.5C charge rate of the same battery is 5A. From previous paragraphs one should notice that capacity and current of a battery is closely related to time: if a battery is discharged at 2C it will be empty after 0.5h. A 0.1C discharge rate will empty the battery in 10h. And of course 1C discharge results in 1h operation. As one can see the C-rate can be used to determine battery runtime without revealing the actual current or capacity of the battery. Table 1 shows some examples of C-rate and related current and charge/discharge time. Additionally C-rate shows how intense the current is in respect to battery capacity. For lead-acid batteries high C-rate discharge will lead to decreased available capacity while very low C-rate discharge will be able to deliver even more charge than the nominal capacity. Often 1C rate does not even correspond to 1h operation.

Battery lifetime is a critical parameter of secondary batteries. Depending on the chemistry battery lifetime is affected by to ageing mechanisms: cyclic ageing and calendar ageing. As the name suggests calendar ageing is related to the absolute age of the battery: as battery ages, its performance will deteriorate – capacity will decrease and internal impedance will increase leading to decreased current capability. The rate of calendar ageing is related to the temperature of the battery. As Arrhenius' equation suggests, the rate of chemical reactions is proportional to the temperature: if battery is operated/stored at elevated temperature, it will deteriorate faster. Additionally the rate of calendar ageing for some battery chemistries is related to the depth-of-discharge. Being fully charged or discharged can increase or decrease battery ageing rate, however it depends on battery chemistry. For example in stand-by applications lead-acid batteries should be fully charged to extend lifetime, while most Li-ion batteries (LIBs) will degrade faster if left fully charged.

The other ageing mechanism – cyclic ageing, is related to the intensity of battery usage. A full battery cycle is a full charge followed by a full discharge. Battery manufacturers in battery datasheets give an estimated cycle life – typically few to several hundreds of cycles. For this cycle number to be true it is of importance to follow a specific charge and discharge test pattern: the manufacturer will specify exact charging and discharging current, exact charging and discharging cut-off criterions and exact rest periods between each charge and discharge as well as the ambient temperature (typically 25°C) at which the battery should be cycled. These cycle life testing parameters are different for each battery model. The estimated cycle life has been reached when the battery capacity has decreased by a certain amount. Unfortunately for different manufacturers and different battery models this end-of-cycle life criterion is different and can be in the range of 85% to 60% from initial nominal capacity – this complicates comparison of different battery models. A key fact is that batteries degrade with each cycle even if the cycle is not full. However this degradation rate and linearity is not the same for all models. For battery with perfect linear degradation each next cycle would be a bit smaller until the capacity eventually decrease from initial 100% to 0%. However, in real life batteries the degradation is nonlinear, especially in the end-of-life (EoL) region. While initial capacity drop per cycle might be linear after reaching the rated cycle life there might be a knee in the cycle life graph after which the degradation happens much faster. This knee point is after the datasheet’s 85% to 60% end-of-cycle life point. Typical battery datasheets does not provide full cycle life graphs from 100% to 0% capacity but only from 100% to the rated end-of-life point which is those 85% to 60%. A key point to note is that the datasheet cycle life number can be easily exceeded if certain actions are taken to decrease degradation. The degradation rate is strongly affected by the temperature, discharge depth (Depth-of-Discharge DoD) and charging level. Deep discharge and overcharging can quickly decrease cycle life. In worst case situations a single full over discharge (well below absolute minimum battery voltage) can lead to immediate battery failure. The same is true for overcharging, especially in the case of LIBs, which can catch fire or explode if overcharged. Even small deep discharges and overcharging will lead to decreased cycle life. High rate discharging and charging can also decrease cycle life if battery temperature is not limited to optimal range. For Li-ion and lead-acid batteries the key action to increase battery lifetime is to limit DoD. For lead-acid batteries it should be as small as possible while fully charged (0% DoD) and fully discharged (100% DoD) regions should be avoided for LIBs.

During battery operation there are two similar but opposite variables which indicate how much charge is left in the battery: depth-of-discharge (DoD) and state-of-charge (SoC). Both of variables are relative indicators and are expresses in percent. As the name suggests DoD indicates how much charge is left in battery in respect to a fully charged battery. A 0% DoD level means that the battery is fully charged while 100% DoD level means that a battery has been fully discharged. DoD can be calculated using (x) where C_dis is discharged capacity and C_nom is the nominal capacity. The SoC is an inverse of DoD: it indicates how much charge is remaining in the battery. A fully charged battery is at 100% SoC while 0% SoC means that the battery is completely discharged. DoD indicator has the benefit to go over 100% - it can happen in situations when a low discharge rate is used and thus battery can deliver more charge than nominal. The SoC indicator in this case would go negative which can cause confusion. In practical use, the SoC is mostly used as indicator to describe the remaining charge of a battery in operation, in essence, it is a fuel gauge indicator. Rarely DoD is expressed not in percentage points but in discharged amp hours (Ah) as shown in expression (x). In this case the user must know the nominal capacity to make sense of the DoD reading.

When selecting a battery for a particular application certain comparison of available models has to be done to make an optimal choice from price and performance perspective. For portable applications gravimetric energy density (also specific energy, gravimetric density) and volumetric energy density (occasionally energy density) are common indicators of battery performance. The gravimetric energy density DE_m can be calculated using (x) where E_bat is the nominal energy and m_bat is the mass of the battery. It is traditionally expressed in Wh/kg units. It shows how much energy can be stored in one kilogram of battery mass. For example, a 300Wh battery weighting 2kg has 150Wh/kg gravimetric energy density. A high-performance LIB has approximately 200Wh/kg specific energy density.

While specific energy describes the relative weight of the battery, volumetric energy density is used to describe the relative volume or size of the battery. A battery with higher volumetric energy density will be smaller at the same rated energy. The volumetric energy density DE_v can be calculated using (x) where E_bat is the nominal energy and Vol. is the volume measured in liters l. The result is expressed in Wh/l units. A battery datasheet most likely will provide the mass however the volume will have to be calculated using provided dimensions. High-performance LIBs can have volumetric energy density of 400Wh/l or even more.

When using both volumetric and gravimetric energy densities one must remember that in most cases these values are for cell-level parts. A fully designed battery pack has more components (interconnections, electronics, thermal management, structural case) which will add to both mass and volume and as a result both energy density performance indicators will be decreased. In most applications the costs of the solution are of great importance. The same is true for battery powered devices. It is occasionally estimated that an EV battery can cost as much as one third of the whole vehicle. The costs of battery can be divided into two parts: cost of cells and cost of auxiliary components. The cost of cells is straightforward – one can use the price set by manufacturer or dealer to compare different cell models. It is more complicated with auxiliary components which are required to make the battery pack/battery energy storage system. These components include battery management electronics, safety devices, contactors, connectors, interconnections, thermal management and the case. In case of EV LIB packs, cell costs make approximately 70% and remaining 30% are the costs of auxiliary components.

One more interesting battery comparison indicator is the lifetime energy throughput. In its simplest version it can be obtained by multiplying nominal capacity with cycle life. The result is an amount of energy which will be delivered by the battery during its lifetime. It must be noted that this calculation does not take into consideration the capacity fading and it assumes that battery will be used only with full cycles. As previously described, the cycle life can be significantly increased if certain operational modes are employed – it would result in much larger total lifetime energy throughput.

For economical perspective, E_lifetime can be divided by the price of the battery to estimate how much energy will be delivered per unit of currency as shown in equation:

For example, a small 300Wh battery costs 150€. Its nominal cycle life is 500 cycles which results in 150000Wh or 150kWh of total discharged energy throughout its lifetime. If this number is divided by the price, then a 1kWh/1€ figure is obtained. It shows that each single euro was converted to 1kWh of discharged energy throughout batteries lifetime. This performance indicator can be used to compare different battery technologies and individual cells which have different cycle life parameters.

As previously described, a battery pack consists of cells and a set of auxiliary components. Both in literature and practice the word “pack” is often omitted as is here as well. For stationary applications there is a term “battery energy storage system” which basically is a battery pack with additional interface converter which takes care of voltage conversion, charging and SoC control. If the interface converter is not integrated in the battery energy storage system, then one could say that it basically is the equivalent of a typical battery pack. Cells of small battery packs (electric scooters, bicycles, drones) are arranged in one group and equipped with basic auxiliary components – bus bars, electromechanical contractors, safety disconnect devices and thermal management. The main reason for less complexity is low cell count relatively small voltage. The voltage of larger battery packs (light- and heavy-duty electric vehicles) are in the range of 300V to 800V. While some sources claim that DC voltage below 120V are low risk, it is safer to assume 48V limit. Such large battery packs are typically composed of several low voltage battery modules which are series connected to produce higher voltage.

Each battery module or a small battery pack consists individual cells. All cells are of the same model and are preferably parameter-matched to provide maximum performance utilization. There are two types of connections which can be used to combine individual cells: series connection and parallel connection. In a series connection cells are connected in a string so that positive pole of one cell is connected to the negative pole of the next cell. The correct polarity is critical. Series connection is used to increase voltage of the pack. A single cell has nominal voltage of 1.2V to 3.8V. Higher voltages can be obtained by series connecting cells. Voltage of a string is the sum of individual cells. n is the number of series connected cells.

In series connection the capacity rating (Ah rating) stays the same as for a single cell – all cells will experience the same current during charging/discharging. However, the total energy of the string will be the sum of individual cells energies. The total energy can be calculated if total voltage and capacity of individual cells is known.

In a parallel connection all positive poles of all cells are connected together, and all negative poles are connected together as well. The correct polarity is of utmost importance as incorrect polarity of a single cell will cause immediate short circuit which in worst case can result in fire and/or explosion. The total voltage of a parallel connection is equal to that of a single cell. Parallel connection affects the total capacity which can be calculated as the sum of combined cells.

As the capacity rating is increased, the C-rate is increased proportionally as well, resulting in higher permissible charge/discharge current levels. Similarly, as in series connection, the total energy of a parallel connection is equal to the sum of individual cells.

One of a battery pack’s description parameters is the cell configuration: how much cells are connected in series and how much in parallel. A thirty cell series connection is described as 30S while ten cell parallel combination is described as 10P. Both parts are typically combined: 30S10P – the battery pack consists of 30 series connected cells and each “cell” is made of 10 actual cells in parallel. This pack contains 300 cells in total. Given cell configuration tells that cells are connected in parallel first (making a larger cell or cell group) and then connected is series to form a single string. An alternative would be 10P30S. In this configuration there would be 10 series strings composed of 30 cells in the first step (series first configuration). In the second step these strings would be connected in parallel to achieve required capacity and C-rate performance.

There exists a mixed cell configuration where in first step smaller series strings are connected in parallel to make a sort of battery module. In second step modules are connected in series to produce finished battery pack. Common 12V lead acid battery has 6 series connected cells (6S1P). One could connect 2 batteries in parallel to double capacity and 2 in series to double voltage: a total of 4 batteries would be used to make a 2S2P configuration – in this case the configuration regards batteries not cells! See figure 3c for clarification.

When designing LIB packs it is a rule of thumb to connect individual cells in parallel first, series second. Such configuration assures that the cells connected in parallel have identical voltage. It is beneficial to parameter-match cells before connecting them in parallel to achieve optimal resulting capacity. During second step, paralleled cell groups are connected in series to build the whole battery pack. A battery management system (mandatory in case of Li-ion) can then be connected to each parallel cell group. If cells were first series connected, then each series string would require individual battery management system due to monitor cell voltages and perform balancing. In parallel-first configuration all parallel cells share the same voltage hence it is simpler to perform monitoring and balancing.

Typical EV LIBs have voltage in the range of 300V to 800V and energy in the range of dozen kWh to 100kWh. It is obvious that such battery packs require a multitude of cells which need to be series and parallel connected to achieve desired performance. Due to reliability, safety and handling issues, such battery packs are composed of smaller parts – battery modules. Modules are connected in series to make a battery pack. The size of the module can depend on several aspects: mechanical integrity – modules are made to be mechanically robust; safety – modules often have voltage (much less than the total battery pack voltage) which can be considered safe, handling – the relatively small size and weight of a module allows easier handling during manufacturing, assembly and repair. While a battery pack consists of series connected modules, the modules themselves can consist of individual cells or cell groups. If individual cells are sufficiently large (tens of Ah) then a module can be made of series connected cells. If cells are smaller (few tens of Ah) then they can be separately connected in parallel to make a cell group or cell pack – several such groups can be connected in series to make a module. However, in some EVs (example: Tesla) the battery pack is composed of modules which are directly made of a large number of small cells (few Ah).

As the battery pack is composed of smaller elements (cells) its shape can be engineered to fit certain requirements. The geometry of individual cells sets the limits of battery pack geometry. Li-ion cells are made in three general shapes: cylindrical, prismatic and pouch. The cylindrical shape is considered to be most mature; it has a set of standardized sizes and some of the most common sizes are produced by a number of manufacturers which permits certain flexibility. A size is described using a five-digit number where the first part describes diameter and second describes length in millimetres. For example, the widespread 18650-size cell has 18mm diameter and 65mm length – these are nominal numbers, some tolerance can be expected, additionally if such a cell is intended for use outside of a battery pack (as a single removable cell), it is internally equipped with a protection circuit, which adds a few mm to the length. Cylindrical cells are manufactured with capacities ranging from a few hundreds of mAh to a few Ah. Both gravimetric and volumetric energy densities are relatively high due to simple manufacturing process and efficient packaging. The outer shell of the cell is made of steel which improves mechanical integrity. From safety perspective, cylindrical cells are equipped with venting mechanism to decrease risk of internal overpressure and explosion and PTC (positive temperature coefficient) resistor devices to limit overcurrent and short circuit current to some extent. The geometry of a cylinder permits good internal pressure durability – cylindrical cells usually does not suffer from swelling. The terminals are similar to those found on consumer AA batteries. The case is the negative pole while the button top is positive pole. Often the button top is replaced with a flat-top terminal which is better suited for spot-welding – a cell-joining technique for electrical connection implementation by welding a metal sheet strip to required cells. For EV applications the relatively small capacity requires a large number of cells (thousands of cells). The cylindrical shape has a disadvantage of poor space utilization at pack level – cells cannot be packed tightly due to their shape, there is always a significant gap. However, this resulting gap can be used for internal wiring and thermal management.

Prismatic shape cells are rectangular, and their dimensions are not standardized which makes it hard-to-impossible to substitute cells in a battery pack with ones from a different manufacturer. The case of a prismatic cell is made of a relatively thick metal (aluminum) or plastic layer which improves mechanical robustness and certain degree of safety. However, the thick casing results in less efficient weight utilization – prismatic cells have the lowest energy density among three cell shapes. From manufacturing perspective, prismatic cells require the most complex production process which results in relatively high final product cost. The regular shape of a rectangle allows easy assembling into a pack without gaps. The manufacturing technology allows production of wide capacity ranges from a few Ah to hundreds of Ah – the largest sizes are particularly suited for stationary energy storage. EVs utilize capacity in the range of few tens of Ah to 100Ah. Prismatic cells are often equipped with threaded terminals to allow connection to large cables or bus-bars – high capacity comes together with high current which in turn requires massive conductors. The prismatic shape is more susceptible to swelling and it is considered normal to have some internal pressure increase during the first months of operation. The elevated pressure results in some swelling which can compromise the mechanical structure of a battery pack.

The third cell shape is pouch. It is similar to prismatic shape - rectangular, but the case is made of plastic-laminated aluminum foil, which makes the casing light weight and small volume. Pouch cells have the most effective packaging – the active battery materials compose well more than 90% of the total mass of the cell. As a result, pouch cells have the highest energy density. The terminals are realized as simple metal strips protruding the cell casing lamination edge, however for single cell use applications, the cell is equipped with a small protection printed circuit board (PCB) which is directly soldered to the cell terminals. Wires are soldered to the output of the PCB and can be seen as the new battery terminals. Same as with prismatic cells, pouch cells do not have a set of standardized sizes. As it currently is the dominant cell shape in portable consumer electronics due to its high energy density performance, it is manufactured in a wide array of sizes and capacities (from several mAh to few Ah) to fit everything from earbuds and smart watches to high performance laptops. As the cell size and capacity increases to 10Ah, pouch cells are used in hobby radio-controlled models and drones. Cells with capacity from 10Ah to 100Ah are used in EVs and HEVs. In EV battery packs pouch cells require extra care as the thin casing material does not provide mechanical strength – the cell can be easily bent or punctured. Pouch cells are typically made thinner than prismatic cells which together with weak mechanical strength contributes to poor internal pressure endurance. It has been observed that pouch cells can swell significantly due to inner gas generation. The exact reasons are not fully defined but it is estimated that excess gas generation happens due to manufacturing issues.

In high performance applications such as EVs, batteries experience both high discharge (acceleration) and charge (fast charging) rates. High C-rates results in elevated internal losses which eventually are converted to heat and as previously discussed, excessive heat is unwanted for both short and long term. In respect to cell shapes, heat dissipation plays a significant role. Battery management system uses temperature sensors to measure temperature of individual cells or temperature at selected locations of the battery pack. A sensor can be placed in direct contact with the cell, but it is important to acknowledge that the sensor will provide a measurement of a surface point. The actual losses originate in internal layers of the cell – only the most outer layer is connected to the casing. The heat of innermost layer must conduct through other layers to reach the surface of the cell. From this, temperature inside the cell will be larger than the surface temperature. Additionally, when heat is passed through the layers, they act as a buffer – inner temperature peaks looks just like small temperature “bumps” on the cell’s surface. In worst case, a sudden high current could quickly overheat center part of a cell while the surface is still cool – if high current persists, the internal temperature rises, and thermal runaway can occur with permanent damage to the cell. Hence it is beneficial to use thinner cells which have shorter distance between center and surface to smoothen temperature gradient throughout the cell. This is one of the reasons why more smaller cells (in series and parallel) are sometimes used in EVs battery packs instead of just large cells connected in series. Each smaller cell can have its own connection to the thermal management while the center plane of a large cell would be located far from the heatsink. Flat and thin pouch cells have the advantage of best heat transfer to an external heatsink as they have the largest surface area (compared to volume). Cylindrical cells have lower surface area per volume thus as the cell size is increased the thermal management issues become important.

The secondary battery market is mainly filled with three battery chemistries: lead-acid nickel-metal hydride and lithium-ion. Among these three types lead-acid technology is the oldest as it was invented in 1859. It is an inherently simple chemistry which throughout years of development has achieved maturity in terms of robustness, manufacturing and recycling. Lead-acid batteries are mostly manufactured in wide selection of standardized prismatic/rectangular packages while some cylindrical models and special purpose shapes exists as well. Notable characteristics are heavy weight (in part due to lead – one of the heaviest metals), low price and high discharge current.

A lead-acid battery cell consists of two lead electrodes submerged in diluted sulfuric acid. Originally both electrodes were made as sheets or plates stacked or rolled together with some isolating material in between to prevent touching of both electrodes and causing short circuit. The number of plates inside a cell is increased to increase the capacity. In general, one could make a cell without any separator material by increasing distance between the electrodes and providing mechanical fixation. State-of-art technology uses various separator materials (plastic, rubber, glass mat) to isolate electrodes or fix electrolyte. Gelled electrolyte can also be used as separator with other beneficial functions. Initially lead-acid batteries were made with liquid electrolyte consisting of diluted sulfuric acid. Such batteries are known as flooded/wet batteries. Additionally, they require vents to release gases which can generate during charging – for this reason flooded lead-acid batteries are sometimes called vented lead-acid batteries. Gas generation is a normal part of flooded battery operation.

When a lead-acid cell is fully charged, the positive electrode consists of lead oxide (PbO₂) while the negative electrode consists of metallic lead (Pb). The concentration of sulfuric acid (H₂SO₄) electrolyte is at the maximum. During discharge (when an external load is applied to the terminals of the cell) the lead (Pb) of the negative electrode reacts with the sulfuric acid (H₂SO₄) and produces lead sulfate (PbSO₄) and free electrons, which in the form of electric current flow through the external load an into the positive electrode. This was the negative half-cell reaction. The positive electrode (PbO₂) reacts with sulfuric acid (H₂SO₄) as well and these additional electrons from other half-cell reaction are included in the reaction which produces lead sulfate (PbSO₄) and water (H₂O) which dilutes the electrolyte. In these reactions sulfuric acid(H₂SO₄) is dissociated into two types of ions: two H⁺ ions and one SO₄^-2 ion. Both reactions can be combined into one as shown in figure 4. It can be seen that during discharge, material of both electrodes is converted into lead sulfate (PbSO₄) and the electrolyte sulfuric acid concentration decreases as the reaction uses negative SO₄^-2 ions of sulfuric acid and produces additional water.

Once the cell is fully discharged both electrodes have completely turned into lead sulfate (PbSO₄) and electrolyte concentration is low. Since both electrodes are made of the same material, there exists an equilibrium of both sides and the cell cannot produce more current. To charge the cell, an external power source (charger) has to be connected to the terminals of the battery to force an opposite current (electron flow from positive to negative terminal). During charging, electrons are taken from the positive electrode and into the negative electrode where addition of electrons releases an SO₄^-2 ion into electrolyte solution and the negative electrode converts back to metallic lead (Pb). Due to the externally applied voltage, positive electrode (PbSO₄) reacts with water to convert back to lead oxide (PbO₂) and produce SO₄^-2 and H⁺ ions. After reactions of both half-cells, electrodes return to the composition of fully charged state (Pb and PbO₂) and sulfuric acid ions increase the concentration of electrolyte as shown in figure 5.

Lead acid batteries are made in two types: flooded/vented and valve regulated. As previously mentioned, the flooded type is historically first. Flooded lead-acid batteries use liquid electrolyte and each cell is equipped with a cap which can be removed to inspect the condition (level and concentration) of electrolyte and if necessary, add distilled water. Water has to be added because battery charging can cause electrolysis of electrolyte water which produce gaseous hydrogen and oxygen. Both gases are released through dedicated vents (typically built into the cap) and water is lost from the electrolyte – it must be replenished for battery to maintain its performance and lifespan. The other type of lead-acid batteries is valve regulated (VRLA – valve regulated lead-acid) or sealed. It is marketed as maintenance-free because there is no need (no option) to add water. This feature is achieved by immobilizing liquid electrolyte using one of two methods. In one case a fiberglass mat is used as the separator. The glass mat absorbs part of the electrolyte and in case of gas generation, it is trapped inside the mat and recombined back to water. This sort of VRLA battery is known as AGM battery (absorbent glass mat). In other case, the electrolyte is immobilized by special additives which makes it into a gel. The gel substance prevents spillage and evaporation of electrolyte and it traps generated gases which are then recombined back to water. This sort of VRLA battery is known simply as gel battery – one must make sure that it actually is a lead-acid gel battery. All VRLA batteries are still equipped with a one-way safety valve to release build-up of gases in overpressure situations.

One of the lead-acid battery advantages is the relatively high discharge current in a wide temperature range of -15°C to +50°C. Discharge rate of general purpose lead-acid batteries is 3-4C while car starter batteries can have 8C rate or more. The voltage of a fully charged lead-acid cell varies from 2.30V to 2.45V depending on application and temperature. This is actually the charging voltage – the voltage will decrease quickly when the charger is removed, and a load is applied. The nominal voltage of a lead-acid cell is 2V however in open circuit it is 2.1V which means that a 12V battery will measure as 12.6V. The discharge curve is considered flat; however, voltage decreases slightly faster as the battery reaches full DoD as shown in figure 6. The discharge cut-off voltage varies with the discharge rate: at high discharge rate the battery should be disconnected when the voltage drops below 1.6V per cell, although batteries with even lower cut-off values exists. If the discharge rate is low, battery must be disconnected at higher voltage: 1.75V to 1.85V per cell depending on the specification of the battery and the application. One must analyze the datasheet of a given battery to use the correct cut-off voltage.

As previously noted, the rated nominal capacity of a battery is given at certain discharge rate. For lead-acid batteries the capacity is often given for 20-hour discharge rate. For example, if a battery is rated 17Ah at 20-hour rate, then the discharge current is just 0.85A. If the discharge current is larger i.e. the rate is higher then logically the discharge time decreases. However, in case of lead-acid batteries the discharge time decrease is not proportional to the discharge rate increase. As the current is increased the discharge time decreases even faster – the available capacity decreases. This phenomenon is described using empirical Peukert’s law given in equation x, where C_nom is the nominal rated capacity at nominal discharge time t_nom (in hours), I is the current at which the battery is being discharged, t is the discharge time at given discharge current I and k is the Peukert constant which for lead acid batteries is in the range from 1 (an ideal battery) to 1.5 depending on the battery technology (flooded, gel, AGM).

For example, if the previous 17Ah (at 20h rate) battery is discharged at 4A then the discharge time is 3.3h and discharged capacity is 4*3.3 = 13.2Ah. If the current is increased further to 34A (2C rate) then the discharge time and capacity drops to 0.27h and 9.2Ah respectively. A Peukert constant value 1.17 was used for given calculations – it corresponds to AGM type. Lead-acid battery datasheets often include graphs and tables which already provide discharge duration at several C-rates together with cut-off conditions.

In basic form, constant current constant voltage (CCCV) charging is used to charge a lead-acid battery. When an empty lead-acid battery is connected to a CCCV charger, it will start charging the battery in constant current (CC) mode – the charger will keep the charging current value constant as set by the user (according to the battery datasheet/specification). During CC charging phase, the voltage of the battery will gradually increase until it will reach another setpoint – the constant voltage (CV) level, which also is set by the user. At this point the charger transitions to CV mode – the battery voltage is kept constant while the current gradually decreases as the SoC of battery increases. The charging is stopped once the current drops below cut-off limit which usually is 3-5% of rated capacity e.g. in case of 10Ah battery charging cut-off is at 0.3-0.5A.

Lead-acid batteries are not suited for what is presently considered fast charging – normal charging will take 6 to 12h depending on the DoD. The initial charging current is well below 1C, typically between 0.4C and 0.15C. Charging with 0.4C is considered “fast charging” and it is used for so called cycle use when battery is fully charged, removed from charger and then significantly discharged. 0.15-0.2C charging is used for trickle/standby/float use when battery is fully charged, left connected to the charger which operates in trickle mode, and then it is periodically discharged to a small extent. When the battery is in cycle use, the charging voltage is higher (14.4V to 14.9V, the voltage is larger at lower temperatures and smaller at high temperatures), when the battery is in trickle use, the charging voltage is lower (13.6V to 13.8V).

The cycle life of lead-acid batteries strongly depends on DoD. When used in cycle mode, at 30% DoD a battery can last as much as 1000 cycles while 100% DoD reduces cycle life to 200 cycles – these numbers vary between battery models. When a battery is used in trickle/float/standby mode, the service life is expressed in years and it is affected by ambient temperature – if operated at 20°C the battery will last 3-5 years, while operating at 40°C will decrease service life to 1-2 years.

Independently from the mode of use, lead-acid batteries should be kept fully charged as often as possible. Deep discharges and prolonged periods at high DoD will result in formation of so called “hard sulfation” of the electrodes which causes capacity and service life decrease. To prevent sulfation, battery should be stored fully charged and periodically recharged to maintain full charge. Trickle charging can be used to keep the battery fully charged – battery is initially charged using CCCV method with cycle or standby current and voltage settings, then charger is set to trickle/standby/float settings to maintain stable SoC. A battery should not be left connected to a cycle mode charger as the elevated voltage will gradually electrolyze the electrolyte which will result in permanent damage and performance decrease of the battery.

Lead-acid batteries have well established areas of applications: SLI, motive, stationary and others. SLI batteries are used in vehicles and SLI stands for starting, lighting and ignition. Lead-acid batteries are especially suited for car engine starting because of the high discharge rate capability. SLI multi-billion € market share is the largest for lead-acid batteries and it is expected to grow in future as more and more vehicles are being produced. Even state-of-art EVs have an SLI battery – there is no more need for ignition but the well-defined 12V system is still used for all auxiliary functions: standby, starting, lighting, HVAC. SLI lead-acid batteries are produced as fixed battery packs with 3, 6 and 12 cells in series in a plastic container with corresponding nominal voltages: 6V, 12V and 24V. 12V lead-acid battery is by far is the most common. It is available with capacities from few tens of Ah to hundreds of Ah. SLI batteries can be both flooded and VRLA.

Motive/traction/propulsion lead-acid batteries are standardized for cyclic use in special electric vehicles: forklifts, automated guided vehicles. These vehicles are used in factories and warehouses – indoors, where internal combustion motors cannot be used due to emission of toxic gases. Battery packs are made of individual flooded cells arranged in a robust case which allows easy maintenance of individual cells. The range of forklift battery nominal voltage starts at 12V and is increased with 12V increments up to 96V depending on the actual application and requirements. The capacity ranges in hundreds to a thousand Ah. Additionally, the heavy weight of lead-acid batteries is beneficial in forklift applications to improve vehicle stability and load-bearing capability. Other motive batteries include cleaning machines, personal mobility vehicles and golf carts. These applications utilize standardized general purpose VRLA and flooded 12V batteries. If higher voltage is required, multiple 12V batteries are connected in series.

As the name implies stationary lead-acid batteries are meant to be installed in a fixed location. These batteries are used for large uninterruptable power supplies (UPS)/backup power supplies and load leveling at utility-scale. In contrast to small UPS devices, the large ones utilize flooded lead-acid cells assembled in battery packs with nominal voltage suitable for direct inverter operation – hundreds of volts. The UPS nature requires the battery to be on standby all the time while it must be able to deep-cycle occasionally as well. The advances of VRLA technology pushes this type to all applications. Single stationary cells have capacities from hundreds to thousands of Ah reaching up to 12000Ah. Small UPS devices use general purpose 12V VRLA batteries with top capacities ranging in few tens of Ah. Often the battery pack is made by using smaller batteries in series-parallel configuration.

A wide variety of shape and capacity (1-100Ah) general purpose VRLA lead-acid batteries are available with 6V and 12V nominal voltages. These batteries are intended for both stationary and mobile applications in various consumer, industrial and professional devices.

Nickel-based secondary batteries have a long history as they were the first real lead-acid rival. The first nickel-based chemistry was invented in the late 19th century and since then several variations have been invented and produced, including: nickel-cadmium, nickel-iron, nickel-zinc, nickel-hydrogen and nickel-metal hydride chemistries. The most notable chemistries are nickel-cadmium (Ni-Cd) which was invented in 1899 and nickel-metal hydride (NiMH) which was invented in 1967, however it was commercialized only in the beginning of 1990s. The first Ni-Cd batteries were made as wet-cells (flooded) and had slightly better performance than lead-acid batteries. They were made of nickel-based positive electrode, cadmium negative electrode and alkaline electrolyte. After a few decades in the 1940s sealed Ni-Cd batteries were produced which afterwards dominated portable electronics market as they had higher energy density than lead-acid (were lighter at the same energy), allowed high rate discharge, fast charging (1C rate) and deep cycling without adverse effects on the cycle life. Notable disadvantages of Ni-Cd batteries are the memory effect, self-discharge and toxicity. The so called memory effect is responsible for capacity drop if a battery is not fully discharged prior to charging. It is generally considered that this effect is not permanent and capacity can be refreshed by performing few full cycles. Ni-Cd batteries have significant self-discharge rate of 10% per month or more, meaning that they cannot be stored and then instantly used – they need to be recharged prior to use. The cadmium metal used in the negative electrode is toxic to humans and other living organisms, hence European Union has banned the use of most Ni-Cd batteries to prevent toxic pollution. Ni-Cd batteries are still being used in special applications and for industrial purposes, especially in the case of large flooded Ni-Cd batteries.

After commercialization in the 1990s, NiMH batteries started to replace Ni-Cd batteries in most applications. In NiMH construction, the positive electrode is a nickel-based material while the negative electrode is made of complex hydrogen-absorbing alloy. The electrolyte remains to be an inorganic alkaline solution. The case is equipped with a safety valve to relieve internal pressure which can form during intense overcharging. All in all, the NiMH is a good substitute for Ni-Cd batteries having just slightly lower discharge rate and being less tolerant to overcharging. NiMH has significantly higher energy density, low internal impedance, flat discharge curve and the chemistry does not contain highly toxic cadmium. One can expect to have a cycle life of at least 500 cycles, while some sources indicate numbers as high as 2000. These numbers heavily depend on the DoD, temperature and charging. In case of small consumer NiMH batteries, sometimes a calendar life of 5 years is quoted while high-performance HEV NiMH batteries have been observed to have usable lifetime of 10 years and more. NiMH batteries suffer from the memory effect similarly as Ni-Cd batteries hence an occasional battery refresh by full discharge can be used to reset capacity. Initial models of NiMH cells had a significantly high self-discharge rate, especially during the first days after charging, however both memory effect and high self-discharge have been improved by chemistry modification and special additives. Some brands have even been labeled as low discharge to indicate that this negative effect has been reduced to an acceptable (similar to other mainstream battery chemistries) level. Even though NiMH batteries are not as robust as Ni-Cd batteries, they are still considered to be tolerant to overcharge and over-discharge. In simple applications a dedicated battery management system is not obligatory to keep battery functional (from safety perspective) as opposed to much more demanding Li-ion chemistries.

NiMH cells have 1.2V nominal voltage, which is considered low and hence requires relatively high number of series connected cells to achieve useful battery pack voltage. The voltage discharge curve is flat with a distinctive voltage drop at the end of discharge (~90% DoD). A fully charged cell is at 1.4V, however the voltage decreases to approximately 1.2V after 15%-20% DoD (figure 7). The voltage curve is affected by temperature and discharge rate. As the temperature is increased, the voltage curve increases. At high continuous discharge rate of 3C, the nominal voltage point is decreased below 1.1V as is the whole voltage curve. Under normal cyclic operation the discharge should be terminated when cell voltage decreases below 1.0V. The minimum discharge voltage is 0.9V – discharge must be terminated at this point especially in multi-cell battery applications. Further discharge (over-discharge) can lead to polarity reversal of individual cells (the ones with smaller capacity) which in turn will increase gas pressure inside the reversed cell resulting to safety-venting. The venting will lead to leakage of electrolyte and rapid deterioration of battery performance. In worst case situation the cell might swell and rupture. NiMH batteries have a wide operational temperature range: most models can be discharged at -10°C to +60°C, although some models are optimized for temperatures as low as -30°C or as high as 85°C. Similarly, as for the most battery chemistries, NiMH available capacity decreases at low temperatures.

NiMH chemistry requires one of the most complicated charging methods. In general, CC charging is used however a multitude of charge termination criterions must/can be considered. NiMH batteries are able to be rapid charged with charging rate up to 1C which results in approximately 90 minute charging time. If a NiMH battery has been over discharged (cell voltage below 0.8V) then it should initially be trickle charged with 0.2-0.3C rate. This trickle charging can continue until 0.8V threshold has been reached and then rapid charging can be initiated. Alternatively, rapid charging could be initiated after 60 minutes. The rapid charging should be terminated after 90 minutes (1st charging cut-off criterion). If low rate charging of 0.1C rate is used then a full charging cycle will last 12-14h. As for all batteries, the voltage of a NiMH battery increases during charging, if the voltage of a cell reaches 1.8V then charging must be switched to trickle mode (at 0.03-0.05C). Full charging cycle should have a timeout at 10-20h (2nd cut-off criterion). Longer charging time will deteriorate battery. If high self-discharge rate batteries need to be kept charged, they can be standby/trickle charged at 0.025C rate. Typical charging voltage limit is 1.5-1.6V. At the end of charging the voltage of a NiMH cell increases reaching the max peak, as the charging continues, the voltage of the cell starts to decrease. When this decrease is 5-10mV charging should be terminated (3rd cut-off criterion) – this is commonly known as ΔV charging method. As NiMH battery approaches full charge, its charge acceptance decreases and charging energy is converted to heat – battery temperature rapidly increases. Charging must be terminated if max charging temperature limit is reached (4th cut-off criterion). Additionally, charging can be terminated if the temperature increase rate becomes higher than 1-2°C/minute (5th cut-off criterion) – this is commonly known as ΔT charging.

Despite the fact that NiMH can tolerate mild overcharging it is generally not recommended. A trickle charge at 0.025C rate most likely (depending on actual battery model) will not cause overcharge however it sill might decrease useful life of the battery. Keeping overcharging minimal is a key factor in prolonging battery life. The optimum charging temperature is in the range of 10 to 30°C. Typically, charging below 0°C is not recommended as it will lead to gas generation and safety vent activation, however advances in NiMH chemistry has resulted in cells that can be rapid charged at temperatures as low as -20°C. At temperatures higher than 40°C charging efficiency decreases, and trickle charging should be used to continue charging. Again, high performance cell can be rapid charged even at 60°C.

The advent and progress of high energy density li-ion batteries have radically decreased the overall market share of nickel-based batteries. The remaining share of Ni-Cd and NiMH batteries has fallen to just a few percent; however, the absolute size of the market is predicted to grow at a steady rate as NiMH rechargeable replaceable batteries remain top choice for consumer applications.

From everyday perspective, the most obvious NiMH application is the replacement of traditional non-rechargeable (primary) cylindrical batteries: alkaline/zinc-carbon AA and AAA batteries. While the both primary battery chemistries provide 1.5V nominal voltage, the NiMH cells rated at 1.2V can be directly used as replacement. However, the smaller nominal voltage can decrease performance in some applications requiring those missing 0.3V, especially in the case of applications where multitude of batteries are series connected to provide higher supply voltage. For example, a six AA primary battery combination will provide 9V while six NiMHs will produce just 7.2V. From capacity perspective, the high-performance NiMH batteries are on par with alkaline batteries. Typical applications include high-drain consumer electronics like cameras, toys and portable music players although low self-discharge NiMH batteries can be used in low-drain remotes, wireless computer peripherals and clocks.

Larger C and sub-C sized NiMH cells have been used as a direct replacement of Ni-Cd batteries in cordless power tools and hobby electronics. The similar performance has granted an initially stable market however recent trends indicate that majority of power tools are now powered using li-ion cells. NiMH cells are still used in industrial/hobby applications where high energy density is not top requirement and lower cost is a key advantage. One industrial example is emergency signs and illumination where NiMH batteries are used and kept on standby using low rate trickle charging.

One of the notoriously remarkable electric vehicles was the EV1 manufactured by General Motors from year 1996 to 1999. This vehicle demonstrated the state-of-art of the day and showed that EVs have a potential in consumer vehicle market. EV1 was on a sort of experimental basis in low quantities (total of couple thousand). Initially, EV1 was manufactured with a lead-acid battery and eventually battery technology was switched to NiMH chemistry using a 26.4kWh pack. Although the production of this vehicle was ceased, it showed that NiMH has potential to be used in EVs. Another notable use of NiMH battery in EVs was the Toyota RAV4-EV which was produced from 1997 to 2014 totaling several thousand vehicles. The first truly successful and well know hybrid vehicle - Toyota Prius used a NiMH battery pack. The production of this vehicle started in 1997 and since then millions of units have been sold. This showed a promising perspective for NiMH battery packs in hybrid and plug-in vehicles. Throughout the first decade of the 21st century a multitude of new hybrid vehicle models with NiMH batteries were introduced however the use of NiMH chemistry in vehicle traction batteries (both EV and HEV) has lately dwindled due to the availability of mass-produced high performance li-ion batteries.

The lithium-based rechargeable batteries were last to enter the market, yet they have evolved quickly, overtaken significant part of the market and the total worth of related technologies is expected to grow in the future. As the name implies, the chemical element lithium (Li) is a key component of every lithium-based battery: primary or secondary, metal or ion variety. Lithium is the lightest metal (lighter than water) with one of the lowest electrode potentials hence it fulfills requirements for a performance battery. However, lithium is highly reactive: it aggressively reacts with water and oxygen from air – both lead-acid and NiMH chemistries uses electrolyte with water hence lithium-based batteries needs a new type of electrolyte. Studies in lithium electrochemistry was already done as early as in the second decade of 20th century. However, the research and development of lithium batteries took off only in 1970s - the first rechargeable lithium battery prototype was demonstrated in 1976. The first commercially available non-rechargeable (primary) lithium battery was already sold in the same decade. It took a couple of decades for the rechargeable battery – it was commercialized in 1991. Initial models used metallic lithium anode (negative electrode) with titanium disulfide cathode (positive electrode) but it was noticed that this construction under certain circumstances and cycling can grow so called dendrites – treelike structures which extend from anode, pierce separator layer and cause an internal short circuit which causes venting with flame, fire or explosion. It was found to be difficult to prevent dendrite growth in lithium metal batteries hence research was shifted to another lithium battery type: lithium-ion battery (LIB) in which there is no metallic lithium and only lithium ions are used transfer and store charge.

Rechargeable LIBs became available in year 1991 with key advantage: higher specific energy density than other available battery chemistries. Gradual development of new cathode materials adjusted LIB technology for most requirements of portable applications from highest energy density smartwatches and cell phones to high current hand tools and long life EV batteries. Additionally, less expensive LIB chemistries were developed for stationary applications. It is estimated that total LIB market is in the 30-billion-euro range and it is expected to increase fourfold in this decade. LIBs constitute approximately 60% of all automotive batteries. Over the 30-year period of commercialization, LIBs have become a dominant battery technology with room for improvement. This technology has changed our lives by enabling personal portable devices and now it is a key-enabler for EVs. In 2019 three researchers were awarded Nobel prize in chemistry for the development of lithium-ion batteries.

Li-ion cell construction is fairly complex if compared to simplistic lead-acid cell. The basic essential components are the same: two electrodes, electrolyte and separator. For basic functionality, electrolyte has to be able to transfer lithium ions at the same time it should not react with those highly reactive ions. While the electrolyte of lead-acid and NiMH cells is rather simplistic inorganic water-based liquid, the electrolyte of li-ion cells is a non-aqueous organic carbonate-based liquid which contains some sort of lithium salt or mixture of salts to provide some free lithium ions for energy transfer. The electrolyte is flammable which adds to the overall safety issues of the LIB technology.

The separator layer has the same basic function – prevent contact of electrodes (internal short circuit) while providing free path for the ion flow. In general, it is made of porous polymer material which can be internally composed of different layers of plastic (polypropylene, polyethylene and others) with additives to provide additional safety by blocking short circuit currents and decreasing flammability. From the cell performance perspective, separator layer is an unwanted element of the cell as it does not contribute to the actual electrochemical reaction. It adds dead weight and volume which in turn decreases volumetric and gravimetric energy density. For this reason, it is desirable to make this layer as thin as possible – some cells can have separator thickness of around a dozen micrometers and constitute just a few percent of the total mass of the cell.

A li-ion cell is made as an ion transfer cell in which both electrodes can accept and store Li ions. During charging/discharging these ions are transferred from one electrode to the other. This operational principle is known as the rocking chair. At this point it must be emphasized that a Li-ion cell can be made using a variety of different electrode materials. There are six most prevalent material combinations which result in six types of Li-ion cells. For description of cell construction, the most popular construction will be used: the negative electrode (anode) is made of graphite (an allotrope of carbon); the positive electrode (cathode) is made of lithium and some other metal/s oxide. Both anode and cathode active materials are selected to provide the required capacity, current and life. Electrodes can have some additives to improve current carrying capability. Additionally, both electrodes are bonded to current collectors (passive material) – high conductivity metal conductors used collect electrons from active material and provide path to the external connection of the cell. In some cell constructions the current collector metal is actually a part of the external connector. Aluminum for positive electrode and copper for negative electrode is a common current collector material choice as both metals have high conductivity, adequate electrochemical stability and are easily available.

The case of a cell is another contributor of passive material. Li-ion cells are being manufactured in all shapes according to the application requirements. In most situations, the internal structure of a cell is made as a jelly roll. First each electrode-current collector combination is made as a sheet roll then both electrodes and separator layer are combined by rolling all layers into one roll. Finally, the electrolyte is filled. Naturally, the resulting roll is of cylindrical shape hence the most effective (from manufacturing perspective) cell shape is cylindrical. In case of pouch/flat and prismatic cell shapes, the jelly roll is pressed and processed into required shape and then encapsulated in casing. Alternatively, a cell can be made of individual material sheets which are stacked together to produce flat cells – this process is more expensive and time consuming than jelly roll process.

The basic operation of li-ion cell is simple: positively charged lithium ions flow from negative electrode to positive during discharge and vice versa during charging. At the same time, during discharge electrons travel through the external circuit from negative electrode to positive and from positive to negative during charging. At fully discharged state the porous structure of negative graphite electrode is empty (without Li ions) and oxidized (without electrons) while the positive electrode is fully reduced to metal (lithium and others) oxide. The situation is reversed when cell is fully charged: negative electrode structure is filled with lithium ions taken from the positive electrode. Corresponding chemical reactions are given below in figure 8.

As previously noted, there are six common Li-ion chemistries. They differ according to the materials used in both electrodes. The widest variety is for the positive electrode (cathode) which can have five compositions: LCO, LMO, NMC, NCA and LFP. These three letters are abbreviations of the main chemical components of the active material. LCO and LMO are among the first commercially available Li-ion chemistries. Their composition is straightforward: L stands for lithium, C stands for cobalt, M stands for manganese and O stands for oxide/oxigen. Therefore, LCO Li ion cell has lithium cobalt oxide (LiCoO₂) cathode (positive electrode) and LMO cell has lithium manganese oxide (LiMn₂O₄ or Li₂MnO₃). cathode. It gets a little bit complicated with NMC and NCA where N is nickel, C is cobalt, M is manganese and A is aluminum. In the names of these two materials lithium (L) and oxide (O) parts are omitted to maintain three-letter format. Therefore, NMC is lithium nickel manganese cobalt oxide (LiNiMnCo₂) and NCA is lithium nickel cobalt aluminum oxide (LiNiCoAlO₂). Finally, in LFP L stands for lithium, F for iron (from Latin: ferrum) and P for phosphate. Therefore, the cathode of LFP chemistry is made of lithium iron phosphate (LiFePO4). In all these five chemistries only the cathode (positive electrode) was the variable. The anode material in all five types remained the same: graphite. The remaining type is LTO where L is for lithium, T is for titanium and O is for oxide/oxygen. Due to some chemical nomenclature rules this material is called lithium titanate (Li₂TiO₃). LTO is used to replace the graphite anode. For some confusion, the cathode of an LTO cell can be made of LMO or NMC material. Additionally, the performance of these types is changing as battery technology is advancing. For example, NMC is a popular type for EVs. It used to have 1:1:1 ratio between nickel, manganese and cobalt hence an extended name was NMC111. Then chemistry was improved to reduce cobalt content (an expensive conflict mineral) and new NMC622 type modification was introduced. It is expected that NMC811 material will be available and become mainstream in near future. The key difference between these variations is the increase in gravimetric energy density.

To summarize, each of six basic types have their specific characteristics: cost, energy density, specific power, safety, life span, temperature range even the voltage. A graphical representation of some different features is given in figure 9. From this, one significant conclusion can be drawn: the name “Li-ion battery” is quite generic – the true performance is revealed when the exact type of chemistry is known. To continue the confusion, a term lithium polymer (Li-poly, LiPo) battery exists. Despite the rumors that Li-poly is some special battery type, it is a type of Li-ion battery which has a sort-of solid electrolyte. In a Li-poly battery, the common liquid electrolyte of a traditional Li-ion battery is replaced with a gel-like electrolyte. In practice, majority of Li-ion cells have some additives and improved separator structure to confine liquid electrolyte thus essentially making Li-poly cells. These cells are mainly made in pouch format. A different variation is the solid-state Li-ion – as the name implies, the electrolyte is made fully solid thus making it possible to produce cells thinner than 1mm. Fully solid-state Li-ion technology is still in research and development stage, however it promises higher charge/discharge rates, longer lifecycles, higher energy density while being safer and less expensive. Most likely all promises will not be carried out but announcements from developer companies indicate that solid-state batteries will become commercially available during this decade. Progress in solid-state technology is intertwined with development of lithium-sulfur (Li-S) battery. Li-S battery could be the next breakthrough in energy density however it heavily relies on functional solid-state technology. A closer future is improved anode materials for existing chemistries. The common graphite anode can be replaced by silicon material which can store significantly more Li ions resulting higher energy density. However, silicon anode cannot provide required cycle life. Both materials are being combined to achieve both features. Advances in carbon materials promise improvements in battery chemistry. One novel carbon allotrope is graphene which excels in high electrical and thermal conductivity – both features can be used to improve performance of traditional graphite-based anodes.

All LIBs are characterized by relatively low self-discharge and no memory effect as opposed to the NiMH chemistry which requires occasional full discharge. In case of LIBs, full discharge is to be avoided to maximize battery lifespan. As for most batteries, the discharge rate affects the voltage of the cell – at high rates the voltage will drop more, in some cases it is beneficial to decrease the cut-off voltage to achieve desired end DoD. In most LIBs the discharge curve (vertical axis represents battery voltage while horizontal axis represents SoC or DoD) is linear with a drop at the final stage of discharge (90-100% DoD) when discharged at low rate. However, as the rate is increased, the drop at high DoD becomes flatter while the voltage drops faster at the opposite end of the curve, at low DoD (Fig. 10). The discharge performance is heavily affected by the temperature of the cell - in figure 9, the 6.6C rate curve does not reach 2.0V cut-off voltage because temperature of the cell has risen to the max limit. The nominal curve is given at 20, 23 or 25°C. At 45°C ambient temperature, the voltage curve of the cell is increased by less than 100mV, hence increased temperature minimally affect voltage under discharge. The situation is different if ambient temperature is decreased. At 0°C the voltage of a cell can a couple hundred mV lower (fig. 11). At negative temperatures the voltage decreases further limiting the discharge rate – if rate is too high the voltage drops below cut-off voltage and discharge should be terminated. This effect is somewhat mitigated if cell is operated at moderate discharge rate – due to internal losses the cell can self-heat and thus improve its performance. For most Li-ion types the available capacity rapidly decreases at low temperature (below -15°C). However, there exists a wide variety of different types and special purpose battery models which are designed to operate at high rates or low temperatures.

It can be said that the type of chemistry plays a critical role in the discharge performance. The most obvious initial difference is the nominal voltage (figure 12). It is commonly assumed that a single Li-ion cell has 3.6V nominal voltage although 3.7V are prevalent as well – these values are for the dominant group of LCO, LMO, NMC and NCA. On top of these two numbers, values around them can exist as well, for example, LG Chem produces 18650-size INR18650MJ1 cell (NMC type) whose datasheet’s nominal voltage is 3.635V. However, 3.6 and 3.7 values are close together and difference is not critically important in most cases.

LCO is rated at 3.6V with 3.0V optimal discharge cut-off. This type is known for its high gravimetric energy density and hence used in space/weight critical applications like mobile phones and laptops. The specific power is poor: discharge rate should not exceed 1C. LCO can be considered as an energy cell – used in applications where high energy is more important than high power. High rates and overcharging can lead to thermal runaway. LCO was one of the initially successful chemistries, but now the high price of cobalt and low safety has made this older type obsolete, putting more focus on NCA and NMC chemistries.

LMO is rated at 3.7V with the same 3.0V optimal discharge cut-off although the minimal voltage can go as low as 2.5V. The energy density is considerably lower than LCO however this type is capable of much larger discharge rates. The recommended discharge rate remains the same – 1C (for max life) but max rate can go as high as 10C or even 30C for brief periods. This high specific power density makes LMO a good choice for power tools and other high rate devices. High current can be achieved due to more stable manganese-based cathode structure – this type is safer than LCO. Thermal runaway would occur at much higher temperature Unfortunately, lower energy density is not the only drawback – cycle rate is lower than that of LCO. Drawbacks have been minimized in NMC chemistry which combines features of both cobalt and manganese, making LMO less relevant.

NMC can be rated at 3.6 to 3.7V depending on the exact materials and proportions of the cathode. The discharge cut-off varies from 2.7 to 3V with 2.5V as the absolute minimum. NMC is considered to be the leading LIB chemistry with ability to produce cells with both high energy and power at good cycle life. Gravimetric energy density is more than 200Wh/kg and discharge rates up to 2C are achievable. For example, the US18650VTC6 cell made by Sony has 3.6V nominal voltage and 3Ah rated capacity at weighting 46.6g. This translates to 10.8Wh energy storage and 231Wh/kg gravimetric energy density. At the same time this cell has 5C (15A) continuous discharge rate. Depending on various additives and construction, NMC is used in power tools, hobby electronics, EVs and industrial applications. More than 60% of all LIBs are NMC and adoption in EVs is more than 50%. The use of manganese improves safety at higher discharge rates however NMC is not considered to be the safest and highest power chemistry.

NCA has 3.6V nominal voltage and cut-off at 2.5V if max discharged capacity is required (100% DoD). It was introduced at the end of 20th century as a replacement for unsafe LCO chemistry. This type is regarded as high energy with good power capability and long life, additionally EV manufacturer Tesla together with battery manufacturer Panasonic has proven that battery packs with price less than 200€ per kWh can be manufactured using cylindrical NCA cells. NCA chemistry does not provide as discharge rate as NMC, the limit is in 2C to 3C range. Additional drawback is the inherently lower safety and easier thermal runaway. These negative effects can be controlled if proper battery and thermal management system is used.

LTO and LFP chemistries significantly differ from the previous two cobalt and manganese chemistries. The nominal voltage of an LFP cell is 3.2V while the cut-off varies from 2.0 to 2.5V depending on the model and mode of operation. If longer life is desired, then cut-off voltage should be increased to 2.8V. The discharge curve is very flat at rapid voltage curves at both ends, hence making it harder to estimate SoC. Despite the low cost of materials, LFP cells are rather expensive because of low gravimetric energy density. The advantages are high power, high safety and long life under specific discharge conditions. Safety includes chemical and thermal stability as well as some tolerance to overcharge and short circuit. For example, A123 Systems manufactures AMP20M1HD-A flat pouch cell which has 19.6Ah at 3.3V nominal voltage (higher than usual 3.2V) at gravimetric energy density of 131Wh/kg which is low if compared to NMC or NCA types. However, this cell can be discharged 363A current which translates to 18C rate. Additionally, the manufacturer claims that this cell model will retain 90% capacity after 3000 cycles at 100% DoD. LFP chemistry is popular in China for EVs, industrial applications and utility level energy storage.

LTO is characterized by even lower nominal voltage ranging from 2.2 to 2.4V. The minimum discharge cut-off is at 1.5V while in some models it is recommended to stop discharging when voltage decreases to 1.85V. The relatively low nominal voltage is the greatest disadvantage of this chemistry. As previously described, energy of a cell can be calculated by multiplying capacity (Ah) with nominal voltage (V). At same capacity and significantly lower voltage the resulting energy and volumetric/gravimetric energy density will be low. This further translates to high initial cost per kWh of the battery pack – the highest among all LIBs. Otherwise, LTO has some significant advantages. Both charging and discharging rates are high: typically quoted discharge rates are up to 10C with 30C pulses. Pulse (10 seconds) current capability of actual high-power optimized models can be as high as 75C. The cycle life is measured in several thousands and if reduced DoD range is used then cycle life can extend to tens of thousands of cycles. Additionally, the operational temperature range is wide and thermal stability is high making LTO the safest Li-ion chemistry. For a practical example, Leclanche manufactures LT34 LTO cell with 34Ah capacity at 2.2V weighting 1080g. Simple calculation yields that gravimetric energy density is just 70Wh/kg – less than high performance NiMH chemistry can provide. However, this cell can be discharged at 6C and 10C in pulses at temperature range from -20 to +55°C. Additionally, at 100% DoD cycling it is rated for 15000 cycles while 80% DoD cycling will extend cycle life to 20000 cycles. Given parameters make LTO suitable for large EVs (bus, tram, train) and stationary energy storage which requires high charge and discharge rates.

In general, the CCCV charging method is used to charge LIBs similarly to lead-acid chemistry. Hence, there are two main charging phases: the faster constant-current phase and the slower constant-voltage phase as shown in figure 13. If a battery is deeply discharged (below minimum voltage) then a pre-charge phase should be introduced before full current CC phase. The pre-charge current should be 10% or less than the nominal charging current (given in the datasheet) of the battery. Once the voltage of the battery is higher than minimal discharge voltage, charger can switch to full current charging in CC phase. In normal operation, pre-charge phase should be omitted as BMS (battery management system) has to prevent deep discharge and associated damage to the LIB. However, if the battery voltage is indicating deep discharge then pre-charge should be carried out to pre-condition both electrodes for effective lithium ion transport. Immediately applying full current (or even worse fast-charge current) to a deeply discharged LIB can result in additional heating (risk of thermal runaway and associated danger) and permanent damage to the electrodes.

Most of the charge to the battery is delivered during the CC charging phase. The controllable parameter is current. The standard charging rate commonly is 0.5C which results in approximately 2-hour 0-100% battery charging (including CV phase). Older Li-ion chemistries were quite sensitive to charging current – higher rates would result in metallic lithium plating on electrodes, electrode expansion (package deformation) and overall performance deterioration. In worst case it would result in thermal runaway and venting with flame. Progress of technology and development of new types (NMC, NCA) have resulted in more robust cells with higher allowable charging rates. Now, faster charging can be achieved by using 1C or even 2C rates. However, fast charging has its limits. In standard charging, most time of charging is spent in CC phase, when battery is charged to 80-90% SoC. CC phase is terminated when charging voltage level is reached and charging transitions to CV phase during which the remaining charge is delivered to the battery. Charging during CV phase happens much slower due to ever decreasing current. When high rate is used in CC phase, the charging voltage limit is reached much faster due to cell heating and resistive drop (seen as voltage increase) similar to that of discharge curve (at high discharge current battery voltage drops, at high charge current voltage steps up). As a result, less charge is transferred to the cell, for example just 60 – 80% or even less. The remaining charge must be delivered in the slower CV phase. The other issue with fast charging is temperature rise. Both resistive losses and ionic conductivity losses produce heat which increase temperature of the battery. When max temperature threshold has been reached, charging current must be decreased hence fast charging transitions to standard charging. This problem can be alleviated if proper thermal management is used for the battery pack. A quality cooling system can keep temperature low (well below max limit, preferably not more that around 30°C) to allow fast charging while avoiding performance degradation. For some battery models, active cooling can allow to increase charging current even higher to achieve faster charging time. In general, it is commonly assumed that EV fast charging (CC phase) can charge battery only to 80% SoC level.

Both LFP and LTO chemistries have some charging advantages. As LFP is thermally more stable it can be charged with 3C rate if proper temperature monitoring is used. The charging performance of some LTO cell models is dramatically different. While typical charging rates can be as high as 6C, cells with 10C and 60C pulse charging capability are available on the market. Some of such cells can be charged to 80% SoC in just 6 minutes.

The CC phase ends and transition to CV phase happens once the voltage of the cell reaches charging voltage limit. For LCO, LMO, NMC and NCA chemistries, the charging voltage is 4.2V. Charging to a higher voltage will result in small addition to the capacity however the cell will degrade faster and the safety risks increase dramatically. Some high-energy optimized cells can be charged to 4.3V however in automotive applications the charging voltage is lowered to improve battery lifespan. Lower charging voltage naturally results in lower max SoC hence it is an easy method to decrease used capacity range. As previously noted, decreased used capacity (never fully charged, never fully discharged) increases cycle life. Additionally, keeping a Li-ion cell at its maximum voltage (same as charging voltage) stresses the internal structure which leads to overall degradation. Lowering max voltage reduces this internal chemical stress and promotes longer calendar life. Again, LFP and LTO max charging voltage is significantly different, same as nominal voltage. Depending on the exact chemistry LFP max charging voltage can be in 3.65 - 4V range. 3.65V is the dominating voltage level while 4V in some datasheets will be given as the absolute maximum level after which damage is imminent. LTO cells can be charged to 2.8 – 3V level – significantly less than other graphite anode-based LIBs.

The charging (SoC increase) speed gradually decreases during CV phase as the current rate declines. In standard charging, CC phase lasts less than 1 hour while CV phase can be in the range of 1 to 2 hours. As previously noted, majority of charge is delivered to the battery during the short CC phase while the long CV phase delivers remaining fraction. If it is required to fully charge a battery the additional step of cell balancing can increase charging time. The charging is terminated when charging current decreases below cut-off limit. This limit traditionally is 10-3% of the 1C rate. Charging cut-off conditions are not always provided by the battery manufacturer, hence there is some engineering freedom. Additionally, charging timeout can be introduced as well. For example, the datasheet of US18650VTC6 cell (manufactured by Sony) states CCCV charging to 4.2V at 2.5A (0.5C) with 2.5h cut-off - a current cut-off limit is not specified. The timeout criterion can be helpful when the battery reaches its end of life. In some cases, the self-discharge/leakage current increases as the battery ages. If the leakage current is higher than current cut-off level, then battery charging current will never decrease below set cut-off and charger will indefinitely continue charging the battery. A timeout can prevent this situation.

Charging is affected by temperature. Cell/battery datasheets provide information about ambient temperature for three situations: discharge, charge and storage. Traditionally one would expect that storage temperature range is the broadest. It is not so in case of LIBs. For short term storage (less than a month) it is the same as discharge operating temperature whose range can be -20°C to +60°C. As the storage time increases, ambient temperature should be kept within -20°C to 25°C range to maintain calendar life. Temperature during charging must be in 0 to 45°C range, preferably below 30°C. Already under 10°C standard charging rate should be decreased to 0.25C. It is generally assumed that LIBs should not be charged if temperature is below 0°C – if temperature is lower, ion mobility is restricted and charging will cause deformation of electrodes, which in turn will degrade performance and safety due to plating of metallic lithium. Both the LIB and the charger should be equipped with temperature monitoring to perform charging only if temperature is within safe operation range. If a battery will be required to be charged at freezing temperatures (an EV in northern countries where the winter temperatures are well below 0°C), then battery pack has to be equipped with thermal management which can provide heating. Of course, the temperature of the LIB will rise on its own during charging due to internal losses – BMS has to prevent temperature rise above operating point. Temperature can be controlled by controlling current or by using active thermal management. It can be noted that charging temperature can vary from model to model and from chemistry to chemistry. Again, LTO excels in operation at low temperatures. An LTO LT34 pouch cell made by Leclanche can be both discharged and charged at temperatures ranging from -20°C to +55°C. Research laboratories are working on improvements for all Li-ion chemistries to allow charging at temperatures below freezing point. There are reports that some LIBs (non-LTO) can be charged at freezing temperatures albeit at very low rates.

LIBs and even single cells require an obligatory BMS which can have a variety of functions. The main task of a BMS is to maintain a safe operation of the battery – the safety of LIBs has always been an issue which requires special care. The functions can be divided in four groups: protection, monitoring, estimation and balancing. Safety essentially is protection. Some cells have some inherent safety features, such as overpressure, short circuit and thermal protection. Overpressure is implemented as a valve which will open when the internal pressure of a cell reaches critical level. Short circuit protection can be made as an internal fuse or a PTC resistive element whose resistance increases if its temperature gets too high – this acts as thermal protection as well. As these three mechanisms are parts of a cell, they do not constitute a BMS which is made using a set of external elements.

In simplest single cell batteries, the BMS (sometimes called cell/battery protection circuit) can be realized as a single printed circuit board which is equipped with a control logic (microcontroller, ASIC or analog), some measurement circuitry and switch elements. The control element uses measurement circuit to measure cell voltage, current and temperature. The measured voltage is processed to provide undervoltage (UV) and overvoltage (OV) protection. UV condition happens when the cell is being deep discharged. OV condition happens when the cell is overcharged. To prevent both of these damaging conditions, the control logic opens switch elements to stop further discharge or charging. Switches can be implemented as integrated elements of an ASIC, as discrete semiconductors (MOSFETs for example) or as electromechanical switches like relays or contactors. The same switches can be used to stop short circuit current (SC) and over current (OC). The SC condition happens when cell current is higher than absolute maximum rating. The OC condition happens when cell current is higher than rated nominal level for prolonged period. For example, some cell can have a pulse/peak current rating of 20A at 10 seconds. If the current is 20A for longer period than 10 seconds, then OC protection should be activated. In some cases, cell manufacturer limits pulse current depending on the cell temperature hence OC protection is often implemented together with thermal protection – high current naturally causes cell’s temperature to rise. When temperature hits some set threshold (over temperature condition (OT)), current must be terminated. Both charging and discharging can be prohibited if temperature is out of safe operation area. This adds under temperature (UT) condition which is particularly important to prevent charging if temperature is below 0°C. When implemented on cell level, these basic protection functions can be realized using small printed circuit boards. For example, the abundant 18650 size cell can be purchased in two variants: unprotected and protected. The unprotected variant is the basic cell with built-in PTC and vent features. The protected variant is equipped with an additional cell protection PCB (18mm in diameter) which is attached to one side of the cell and the whole package is covered with plastic insulation. The resultant cell is a few millimeters longer as opposed to the unprotected one which is 65mm long. A small protection board is incorporated in even the smallest pouch cells to provide at least minimal protection. In simplest form, the protection board can lack logic circuits and switches. A resettable fuse (for example PPTC) could be used instead.

The same protection features are implemented on battery pack level as well. Naturally, the complexity of the BMS increases with number of cells. In a multi-cell pack, parameter monitoring can be separated as a distinctive function. Designated front-end integrated circuits are being manufactured to fit most battery pack requirements. These ICs perform individual cell voltage measurements, pack voltage measurements, pack current measurements and temperature measurements. Some safety actions can be implemented in the monitoring ICs while others are performed by central controller. Front-end monitoring ICs are equipped with some sort of communication protocols to transfer obtained values to higher level controllers. In large battery packs (such as EVs) the battery is split in modules and each module can have its own module management board which transfers individual cell data to the central BMS board where it is processed, and appropriate actions are taken. Some new variables are generated at the pack level: max and min cell voltage, pack voltage, max and min temperature. These values can be further used not only for basic protection functions but for estimation of various performance indicators and cell balancing. Current monitoring can have its own board or at least IC. Shunt or Hall-effect sensors are used to measure instant values which can be used locally for fast short circuit protection or sent to central controller for advanced processing. For large high-performance packs, thermal management becomes and important issue as proper temperature conditions can greatly expand life span of LIBs. Thermal management system can be a part of overall BMS. Battery packs can be actively cooled (or heated) – temperature of coolant medium and its flow must be monitored as well. BMS monitoring functions might include data logging of all mentioned measured parameters and additional ones like total cycles, max and min discharge levels, total delivered energy and other time and charge related variables.

The central BMS controller uses data from monitoring circuits to implement safety and protection functions. Additionally, data is used to calculate and estimate various performance indicators which can be further used to govern the pack or sent to higher level controllers and user interface. From overall vehicle system perspective, SoC estimation is one of the most important functions as it provides information about the remaining available proportion of charge. There are various SoC estimation algorithms depending on battery chemistry and application. Traditional input parameters are current which is being integrated over time (known as coulomb counting) and voltage of the pack which in turn depends on the instantaneous current value and temperature. As the battery ages, the inner parameter values change and SoC estimator must adapt to those changes. Additional information can be obtained from battery electrochemical impedance spectroscopy (EIS), which in some cases can provide direct information about SoC and in others it can aid to determine other parameters. It must be noted that SoC provides estimate in percent of nominal capacity and not the actual available energy. When the battery is new, 100% SoC indicates that full rated energy is available, however as the battery ages, the capacity decreases thus after some years of operation the SoC of a fully charged battery will be 100% but the actual energy might be just 80% of what it was when the battery was new. This brings to another battery performance indicator: the state-of-health (SoH). In simplest form, SoH indicates how much the capacity of the battery has degraded. It can be estimated by dividing actual capacity by nominal rated capacity. The result is expressed in percent. A 100% SoH indicates that battery is new – this number will gradually decrease as the battery ages. SoH can be defined differently: it can represent the remaining useful life (RUL) of the battery. In this case the estimation becomes a lot more complex as it takes into consideration the impedance change (using EIS), the cycle life (in form of cycle counter) and other parameters like self-discharge. The result can be expressed in percent, remaining total storable energy or even remaining days before failure. Another parameter to estimate is the state-of-power (SoP) or simply available power. The available power depends on the ambient temperature, temperature of pack and SoC. It indicates how much power for how long time can be discharged. This parameter is important is systems where future activity (high discharge rates) can be and has to planned.

These previously described BMS function groups (protection, monitoring, estimation) to some extent are common for all high-performance/high-cost/high-reliability applications and battery chemistries. The last function – balancing, is not obligatory for most chemistries (although it can be used in all), however it is essential for LIBs. Lead-acid and NiMH chemistry cells can be balanced by trickle charging; however, it is not allowed for Li-ion chemistries, hence external circuits must be introduced to provide balancing. The balancing function is used to keep all cells of a battery at the same charge level although typically, balancing keeps cells at the same voltage level. There are two reasons for cell mismatch. The first one is that all cells are not created equal – manufacturing differences affect capacities, leakage and other parameters of individual cells. The other cause of misbalance stems from usage and ageing – temperature gradient, interconnection structure and just plain ageing can cause cells to develop different capacities, OCV curves and leakage over time. When all cells of a battery pack are perfectly balanced, the available capacity is maximal. However, if one cell is at lower SoC, it decreases the SoC of the whole pack – it will be the first cell to reach discharge cut-off threshold and hence the battery is rendered empty although other cells still have usable charge. This is the case where the weakest link determines the strength of the chain. Additionally, when a misbalanced pack is charged and a single cell is misbalanced at higher SoC, it will reach full voltage faster. If charging is continued, this cell will be overcharged with all damaging consequences. Hence charging must be stopped when any of the cells of a battery pack reach full voltage.

A variety of balancing topologies can be used to equalize cell voltages/SoC levels to maintain optimal battery performance. The simplest and most widespread is the passive balancing. The topology consists of a switch (for example MOSFET) which connects a resistor across a cell whose voltage is too high (more than other cells). The resistor is referenced to as shunt or bleeder resistor as its removes excessive charge. The operation of the switch is controlled by BMS balancing controller. Commonly, this passive balancing (also know as resistor balancing, shunting balancing) is done only during the end phase of the charging. When first cell reaches full voltage, the charging current is decreased to prevent overcharging the full voltage cell while still charging remaining cells. The balancing process can take a long time as shunting resistor usually provide just a few mA of discharging current. Hence it can be continued after the charger is disconnected from the pack. Balancing at the full SoC is known as top balancing. Alternatively, balancing can be done at discharge: all cells are fully discharged using balancing resistors to achieve equal SoC. Passive balancing while being simple, easy to implement and cost effective, has its drawbacks: the balancing rate is low and can take hours to achieve good balance (depending on the level of misbalance); it wastes energy – all excessive charge is dissipated as heat in the shunting resistors. An alternative is to use active balancing. A variety of topologies have been developed which can transfer charge from full cells to ones with less charge thus preserving valuable energy. Active balancing can be used at all times as opposed to passive which usually is used only during charging. It can be designed to provide high balancing current to deal with high mismatch. However, the main drawback is complexity which adds to cost and reliability issues. Passive balancing is sufficient in most situations, especially if cells of a battery pack are sorted and matched prior to pack assembly – only miniscule balancing is required during charging for most of battery life. Furthermore, battery chemistries and manufacturing process is being continuously improved to minimize cell parameter dispersion (both of fresh and aged cells) and increase overcharge tolerance thus reducing requirements of balancing circuits.

BMS and its functions can extend further. As mentioned, thermal management can be a part of BMS, especially if active temperature control is used: forced air or liquid cooling/heating. In EVs charging is controlled by the BMS as it has information about charging voltage, current and can provide temperature and safety control. BMS has communication interface to the main vehicle control system. Some sort of charger is usually implemented in an EV and in some cases the charger is a part of the battery pack. EV batteries are equipped with fuses and set of contactors: for work current and precharge. Smaller batteries can have some human machine interface (HMI), for example a set of LEDs, to indicate remaining charge. All of these features are controlled by the BMS.

Fuel cells (FC) are devices somewhat like batteries. Their purpose is to provide electrical energy by converting chemical energy. Same as batteries they are electrochemical devices. The main difference is that fuel cells uses some sort of chemical compound (fuel) which is supplied to the cell to produce electricity by controlled electrochemical redox reaction. The fuel and oxidizing material (oxygen from air) is consumed during the process. On the other hand, battery already had all components embedded in a closed package and no material was consumed during charging/discharging. Fuel cell technology has a long history as it was invented in 19th century. Since then various configurations have been developed and implemented in stationary or portable applications ranging from few Watts of power to mega-watt systems. The most notable fuel cell technology is the proton exchange membrane fuel cell (PEM FC). The basic elements of a fuel cell are anode, cathode and electrolyte. A PEM layer contains electrolyte and separates anode from cathode. Fuel is delivered to the anode side while oxygen is delivered to the cathode side. Popular fuels are hydrogen and methanol. As an electrochemical reaction takes place, protons from fuel are transferred through the PEM to the cathode side where waste is produced: water in case of hydrogen fuel and CO2 if methanol FC is used. As usual, electrical load is connected to anode and cathode to deliver electrical energy. The reaction is unidirectional – PEM FC delivers energy and it cannot be charged.

Several FC cells are stacked together to increase current and/or voltage thus making a FC stack similarly as in the case of battery packs. The nominal voltage of a single PEM FC cell is approximately 1V and it is load dependent. A hydrogen PEM FC is considered compact power source (high specific energy and energy density) and it can be refueled quickly – one just needs to fill the hydrogen container – a process which takes just a couple minutes. These features make hydrogen FCs a good candidate for powering EVs and other mobile applications. However, widespread adaptation has not been successful so far due to some disadvantages. PEM FCs have high initial costs, in part due to expensive catalyst materials, additionally as FCs age, their voltage and power output decreases hence they have a limited life span (operating hours). FCs are not as efficient as LIBs. Common efficiencies are in 50 to 60% range which means that significant amount of heat will be generated during power production – a cooling system like the one of common ICE vehicles is required as the temperature operational range of FCs is limited. The high cost and limited power density results in requirement of additional energy storage element – a battery or supercapacitor. FCs have an optimal power output level which usually is lower than the power required to accelerate a vehicle. During constant speed driving there is supplementary power which can be used to charge a battery which in turn can be used to provide the required power boost during acceleration. A battery is beneficial for regenerative braking – FC cannot be used to store braking energy. A key issue for hydrogen FC adoption is the lack of refueling infrastructure. Hydrogen gas is extremely flammable, it can diffuse in and through metals and it can cause metal embrittlement hence manufacturing, handling and storing hydrogen requires additional care. The final problem is the source of hydrogen. The amount of hydrogen in atmosphere is negligible hence it must be manufactured. It can be produced by electrolysis of water however this process is inefficient even further decreasing the total FC technology full-cycle efficiency. Currently, majority of hydrogen is produced by reforming fossil fuel – not a sustainable solution. Despite these drawbacks, hydrogen FC technology has been and is used in some commercial EV products. There are a few available automobile models and several public transport buses. The later has shown good performance as buses have stable cyclic usage and plenty of space for FC and hydrogen storage tank installation.

In simplistic terms, supercapacitors (SC) are capacitors with extremely high capacity. In fact, they use special physical effects (electrochemical pseudocapacitance and/or electrostatic double-layer capacitance) to provide capacity. Depending on brand-names and physical effects, supercapacitors are also called boost capacitors, ultracapacitors, pseudocapacitors and electrostatic double-layer capacitors (EDLC). One must not confuse SCs with common high capacity aluminum electrolytic capacitors which are made with rated voltages from few to hundreds of volts. The rated voltage of a single SC cell is in the range of 2.1 to 3V. While majority of SCs are available as single cells, commercial SC batteries (series/parallel) with voltages higher than 100V are available on the market. The capacity of single SCs ranges from hundreds of millifarads to a few kilofarads – they extend capacitor capacity range as the largest electrolytic capacitors are just around 1F in capacity. If compared to LIBs, SC main advantage is the high specific power density (up to 14kW/kg) and vast cycle life (1000000 cycles). However, they have not replaced LIBs or other batteries due to relatively minuscule specific energy (7.4Wh/kg for 3400F capacitor) which makes them inappropriate for bulk energy storage. SCs can be used to improve the power capability of battery systems, especially in applications where the bulk energy storage is a relatively low capacity battery (with low power) as in hybrid vehicles and fuel cell vehicles. SCs can be successfully implemented in applications requiring regenerative braking: the power handling capability of SCs is perfect for absorbing high power pulses of braking while the stored energy can be used for the following acceleration also requiring high power.

SC technology is evolving to improve the overall performance. Hybrid capacitors have been developed – they use both SC and Li-ion technology. The result is so called lithium ion capacitor – as name suggests, it is more like a capacitor with some features of the Li-ion battery. The main advantage is elevated voltage: 3.8V rated voltage increases the specific energy narrowing the gap between SCs and LIBs. Future improvements using graphene materials could increase performance of SC, the same is true for LIBs.

To conclude this chapter see figure 14 - a Ragone plot which is an effective tool to graphically compare gravimetric energy density (specific energy) and gravimetric power density (specific power) of various energy storage elements. The lowest performance is at the bottom left corner while the highest performance is at the top right corner - a spot to be taken by future technologies. Given figure represents generalized performance - higher performing application specific technologies exist and are under continuous development. As can be seen fuel cell technology can provide the highest specific energy while capacitors can provide the highest specific power. However, Li-ion technology with its high specific energy and good specific power is the right choice for most mobile/portable applications.