From the 1970's through the 1990's there was a buildup of tension surrounding the promise of better battery technologies that would make electric cars and other battery storage systems practical. Lead acid batteries had never worked very well, and were extremely bulky and heavy. Nickel based batteries seemed to hold promise, but the nickel cadmium batteries were an expensive and environmentally unfriendly distraction. Finally in the late 1990's and especially in the first years of the 21st century lithium based batteries came onto the scene as the realization of the dream of a functional battery system. For an account of lithium ion batteries onboard Eva see the "LFP House Bank" section of
Failure Fiasco.
Advantages of Lithium Ion Batteries
Types of Lithium Ion Batteries
Management of Lithium Ion Batteries
Thermal Management
Cell Balancing
Nickle Based Storage Batteries
Generally speaking lithium ion batteries are less than half as heavy as other types of batteries, and also perform radically better. Very small amounts of voltage sag during heavy discharge and the ability to be charged quite fast without overheating mean that lithium ion batteries can do big electrical storage jobs. Another way to describe the higher efficiency of lithium ion batteries is to say that the columbic efficiency and the voltage efficiency are higher than for lead acid batteries. The lower columbic efficiency of lead acid batteries means that 115 or 120% as many amp hours have to be put back in during charging as are taken out during discharging. For lithium ion batteries usually less than 105% as many amp hours have to be put back in as are taken out during discharging. The lower voltage efficiency of lead acid cells is due to them supplying only about 1.8V to 2.0V during discharge yet requiring about 2.3 to 2.4V during charging, which is a voltage efficiency of around 80 to 85%. Lithium ion batteries on the other hand typically have voltage efficiencies of more like 90 to 95%. What this all adds up to is that lead acid batteries are less than 70% efficient, where lithium ion batteries are typically more than 90% efficient. When lead acid batteries are run harder to supply large currents the voltage efficiency drops off dramatically, lithium ion batteries on the other hand can support very large discharge loads without the voltage efficiency plummeting. The fact that lithium ion batteries generally don't mind being stored in partial states of charge also means that battery systems can be more flexible and more effective compared to lead acid batteries.
Lithium ion batteries often get lumped in together as just one type of battery, but the reality is that there are three distinct types of lithium ion batteries that have quite different characteristics. The most important distinction among lithium ion batteries is between the 3.7V cells and the 3.2V cells. The 3.7V cells are the lithium cobalt oxide (LiCo) cells and the lithium manganese oxide (LiMnO) cells. Sometimes the lithium manganese oxide cells are called 3.6V cells because they do have slightly more voltage sag during heavy discharge than the lithium cobalt oxide cells. Even though both of the 3.7V cell chemistries have very similar charge and discharge characteristics there are some significant differences. The lithium cobalt oxide cells are able to attain the highest energy densities of 190Whr/kg or more, where the lithium manganese oxide cells are only able to do about 140Whr/kg. The differences don't end at the maximum attainable energy density either. The lithium manganese oxide cells are both cheaper and much more stable, which has made them increasingly popular in recent years. It was the lithium cobalt oxide cells from the late 1990's that were able to so spectacularly fail, bursting into flames or even in some rare cases exploding.
Ultimately though both the lithium cobalt oxide and lithium manganese oxide 3.7V cells are rather similar, and it is the 3.2V cells that are radically different. The 3.2V lithium ion cells are the lithium ferro phosphate (LiFePO4) chemistry, and are in many ways far superior to the 3.7V cells. The major detractor of LiFePO4 cells is the fact that they have considerably lower energy densities than lithium cobalt oxide cells, and even the lithium manganese oxide cells usually attain somewhat higher energy densities. Lithium ferro phosphate cells can attain energy densities as high as 130Whr/kg, but about 100Whr/kg is typical for heavy duty cells. The advantages of LiFePO4 cells are however significant. One really big difference is that only about a tenth as much lithium is required for LiFePO4 batteries compared to the 3.7V chemistries (based on 2009 restrictions on the transport of lithium ion batteries by commercial airlines). This is really very significant since lithium is an expensive, dangerous and potentially environmentally hazardous material to work with. The much lower lithium requirement for LiFePO4 cells means that they should be a whole lot cheaper. Although LiFePO4 cells have at some times been available at significantly lower prices than other lithium ion chemistries, they usually have sold for very nearly just as much. One of the reasons for this is that LiFePO4 are often considered the premium cells because they have so many performance advantages over other lithium ion chemistries. The main thing that LiFePO4 cells do a lot better than lithium cobalt oxide or lithium manganese oxide is hold up to heavy use over long periods of time.
The maximum cycle life of lithium ferro phosphate batteries is usually listed at no less than 2000 deep cycles, where other lithium ion batteries are usually listed as having a maximum cycle life of just 1000 deep cycles. Lithium manganese oxide cells certainly can do 2000 or more deep cycles if they are well managed, but by the same token lithium ferro phosphate cells have the potential to do many times more cycles than this if they are well managed. The really big difference that is often observed though is that lithium ferro phosphate cells will do a whole lot of cycles even if they are run extremely hard and poorly managed, where lithium manganese oxide cells often die very quickly when they are run a bit too hard or poorly managed.
One of the big advantages of lithium ferro phosphate cells is that they are more efficient under both charge and discharge, meaning that they heat up less. Probably closely related to this higher efficiency is the fact that lithium ferro phosphate cells are usually listed as being able to support higher maximum sustained discharge currents as well as higher sustained charging currents than other lithium ion cells. There seems to be essentially no limit to how rapidly lithium ferro phosphate cells can be charged, so long as they are not allowed to overheat, whereas lithium manganese oxide cells usually require much lower charging currents to last well.
The main thing that all lithium ion batteries require to last well is to be prevented from being overcharged. This is especially true with the 3.7V cells, which can be quickly destroyed by being held at a high charge voltage for long periods of time. The lithium ferro phosphate cells are much more resistant to damage from being held at a high charge voltage, but they too need to be prevented from being held at a high charge voltage if they are to last as well as they can. The basic idea is that lithium ion cells need to be charged up to slightly less than 100% capacity if they are to do a large number of cycles. The usual recommendation is a 95% state of charge, but this is a very arbitrary number. The important thing is not overcharging the cells, and if this can be accomplished at a 99% state of charge then a 99% state of charge is perfectly acceptable. If on the other hand preventing overcharging can only be attained by aiming for a 90% state of charge then this is what is required.
What is tricky about attaining a 99% state of charge is that just what the voltage and current conditions at that 99% state of charge are is different for different models of cells, and also changes both with temperature and with the age of the cell. A 95% state of charge is easier, because it leaves much more wiggle room. Just about any 3.7V cell can be charged up to 4.0V at 0.5C until the amperage drops off to 0.3C at normal temperatures. If the cells are charged at higher rates though they heat up considerably, meaning that charging must be stopped before the amperage drops off so far.
Lithium ferro phosphate cells are a bit more forgiving in that they don't mind being held at the charge voltage for a short period of time to attain a full state of charge. Ideally lithium ferro phosphate batteries should not be regularly charged beyond 95 or 99% capacity either, but this also is a bit easier to attain since even at higher temperatures the charge current drops off as a 100% state of charge is approached. The fact that lithium ferro phosphate batteries heat up less when being charged at high rates also makes hitting the 95 to 99% state of charge easier. Just what the charge voltage should be for lithium ferro phosphate batteries is however not entirely clear. The usual recommendation is 3.65V per cell, but this is in fact probably a bit too high for most applications. Down at 3.57V per cell lithium ferro phosphate batteries charge up quite quickly, and easily attain a full state of charge even in cold conditions. If a bit longer taper charging period is acceptable really all that is required to get lithium ferro phosphate batteries fully charged is about 3.52V per cell. Even in quite hot conditions though the 3.52V per cell charge voltage certainly does add some taper charging time, perhaps a half an hour or so. Down at 3.47V per cell long taper charging periods of several hours can be required, but again high states of charge are still attainable.
It is usually said that lithium ferro phosphate cells do not require that the charge voltage be temperature compensated, and this does seem to be true. With the same 3.52 to 3.57V per cell lithium ferro phosphate batteries charge up quite well over a wide range of ambient temperatures. When the cells are cold though they do require longer taper charging periods. If the goal is to rapidly charge LiFePO4 batteries in cold conditions then a high charge current is appropriate. The high charge current gets more energy into the cells quickly, and it also heats the cells up so that by the time the taper charging period at the end is reached the cells are warm enough that only a few minutes of taper charging will be required. Since lithium ion batteries don't mind being stored in a partial state of charge there is usually no great need to do any taper charging. Once a 90% state of charge has been attained charging can be ended.
Even though most lithium ion cells are very resistant to damage from deep discharges there is reason to limit the depth of discharge. The maximum cycle life can be considerably longer if the last bit of the capacity is not used. Stopping discharge before the cell is fully dead is fairly straight forward when high discharge currents are used. Under a high discharge current the voltage drops off dramatically towards the end of the capacity. Under lower discharge currents it is a bit trickier to limit the discharge before 100% capacity has been used, but there is still a very noticeable voltage shoulder towards the end of the capacity. For all types of batteries by far the best way to limit maximum discharge is with the use of an amp hour counter so that discharge at low currents can be ended before the voltage shoulder is reached. Lead acid batteries can be made considerably cheaper, lighter and better performing with the use of stronger electrolyte and thinner lead plates if discharge can be ended before the voltage shoulder has been reached. It is not clear whether the lithium ion chemistry cells can similarly be made cheaper and lighter by ending discharge before the voltage shoulder has been reached. Traditionally lithium ion cells have been nearly totally resistant to damage or degradation from full discharges, but it is not clear if this is an inherent difference in the lithium ion chemistry or if it is just a case of heavy and expensive cells being produced specifically for robust over discharge protection.
Aside from preventing overcharging and over discharging the other main concern for lithium ion battery management is preventing over temperature conditions. Lithium ion batteries can support very substantial 3C or even 10C sustained discharges, but these high currents cause the cells to heat up rapidly. Just how long high currents can be used depends on the size and shape of the cells and what type of cooling system is used. Because lithium ferro phosphate cells are more efficient and heat up less, larger size cells can be used with no active cooling system in many high power applications. The 3.7V cells on the other hand heat up much more, and require smaller size cells with better active cooling systems. Because lithium ferro phosphate cells can support higher discharge currents than the 3.7V cells it remains true though that high power lithium ion battery systems do require small cell sizes and active cooling systems. Small 3Ahr cylindrical lithium ferro phosphate cells in open air can do 10C discharge for most of their capacity. At the end of this approximately four minute discharge the 3Ahr cells are however too hot to handle. Keeping cells from getting this hot allows them to last much better. Just how hot the external case of a cell can safely get depends on the size and shape of the cell. Thinner cells can safely run with higher case temperatures, where thicker cells end up much hotter in the middle and may already have been damaged by the time the case temperature seems excessively high. For this reason high power battery systems are best made up of thin cells with an active cooling system that forces air or water between the cells.
Air cooled cells tend to work better for most applications for two reasons. One obviously is that water cooling requires more substantial isolation of the cells and wiring from the water. This generally means heavier sealed aluminum separators between the cells for the water to flow through. The other reason that air cooling works better is that the air has a much lower specific heat, meaning that the cells can more easily heat up to operating temperature with the cooling blower turned off. For absolute highest possible sustained power output though water cooling does work better simply because the maximum rate of heat transfer from aluminum to water is higher than for aluminum to air. The higher specific heat capacity of water also means that the passages between the cells can be smaller, further increasing the maximum power density of the finished battery system.
The key to any successful active cooling system for lithium ion cells is a blower or pump that comes on not when the cells overheat, but rather at the beginning of a sustained high current charge or discharge. Likewise the blower or pump needs to run on after the high current charge or discharge, but just how long it needs to run depends on how soon another high current charge or discharge event is expected. The general goal is to keep the cells warm enough to perform at their best, but also to draw the temperature down as far as possible during and after sustained high current events.
A battery system of any chemistry can potentially suffer from an out of balance condition, but some battery chemistries inherently self balance much better than others do. Lead acid batteries self balance very easily since they have to be so substantially overcharged on every cycle to last for a long time. Lithium ion batteries on the other hand generally do not self balance very well, and out of balance problems can occur. The larger the number of cells in series the less likely the battery is to stay sufficiently in balance, and also the larger the potential problems will be in the event of an out of balance condition developing. Lithium ferro phosphate cells self balance better than other lithium ion cells, and it is generally believed that four cell LiFePO4 batteries do not need any active cell balancing system. This certainly does seem to be true, as four cell 13V LiFePo4 batteries nearly always do seem to stay well in balance. Three cell 11V lithium ion batteries also appear to stay well in balance with no active cell balancing system, but this is considered a bit more risky. A big reason that running the 3.7V cells without active balancing circuitry is considered poor practice is that it is the 3.7V cells that sometimes so spectacularly die a fiery death when they are radically overcharged. The lithium ferro phosphate cells on the other hand normally don't do anything more spectacular than simply stop working if an out of balance condition is encountered. Of course lithium ferro phosphate cells that are being run very hard can get quite hot, and if some sort of a management system failure causes an over temperature condition there is always the possibility of nearby plastics melting or catching fire.
There are two basic things that cell balancing systems seek to avoid, that is over voltage conditions or under voltage conditions of individual cells. For lithium ferro phosphate batteries the big thing about individual cell management is simply preventing cell reversing if the battery is fully discharged. These individual cell management systems for lithium ferro phosphate batteries normally allow individual cell voltages to drop down as low as 2.5 or even 2.0V before pulling the plug. This does not do much to prevent the battery from being fully discharged, as there is really nothing left in the cells beyond 2.0V. All it does is prevent the cell reversing that can so quickly kill lithium ferro phosphate cells.
Cell balancing to prevent individual cells from charging at higher voltages than others is much trickier, and has the potential to do much more harm than good. The problem with "top balancing", as regulating individual cell charging voltages is often called, is that it can actually cause just the out of balance condition that it is supposedly attempting to prevent.
A good individual cell management system for lithium ion batteries does not actually do any active cell balancing, but rather simply prevents catastrophic failure in the event that an out of balance condition does develop. During charging the individual cell voltages must be allowed to climb to slightly higher levels so that the battery can self balance. If all individual cell voltages are capped at some low level then no self balancing can take place, and the battery will eventually get out of balance. A functional individual cell management system for lithium ferro phosphate cells will allow individual cell voltages to climb as high as is required during normal charging, but then will cap the maximum individual cell voltages at perhaps 4.0 or 4.2V to prevent heavy releases of electrolyte in the event of battery failure. During the life of the battery these higher maximum individual cell voltages would never be encountered, and it is only in the event of a spectacular failure of the battery system that the high voltage cutouts would come into play. Essentially the high voltage individual cell limits are only used as an added measure of safety in the event of battery failure. An additional sophistication would be individual cell temperature limits. If any one cell gets too hot during charging then the charging amperage is backed off to prevent further over temperature problems.
The most important feature of any individual cell management system is the individual cell low voltage cutout. This feature of an individual cell management system might actually be used many times during the life of a battery system. Of course for maximum cycle life it is best to end discharge before these low voltage limits are reached, but dropping down to the low voltage limit is only a bit hard on the cells. The way the low voltage limits work is simply to use the weakest cell to end the discharge. When the weakest cell reaches the minimum voltage the system cuts out. In this way a battery can continue to be used for a long time at the end of it's life even if it's capacity is severely diminished and the balance of the cells is very poor. In the absence of an individual cell low voltage cutout the end of the life of a battery tends to extremely abrupt, with total failure of the weakest cell occurring as soon as cell reversing begins to occur.
Individual cell management systems for the 3.7V cells would be similar in nature, but the voltages at which they operate are a bit trickier to set because of the finicky nature of the more reactive cell chemistries. The main problem with capping the individual cell charge voltage for the higher capacity lithium cobalt oxide batteries is that they tend to start doing spectacular things at only very slightly higher voltages than the normal charging voltage. Essentially the same 4.2V maximum limit applies to all lithium ion chemistries, but this is only five percent higher than the normal charging voltage for 3.7V cells as opposed to being seventeen percent higher than the normal charging voltage for the 3.2V lithium ferro phosphate cells. Individual cell temperature limits are a particularly important part of successful management of the highest capacity 3.7V cell based lithium ion battery systems.
Nickel based batteries don't attain quite the same huge energy densities that lithium ion batteries do, but nickel based batteries can have many of the same longevity and usability advantages over lead acid batteries. Somewhat ironically nickel iron (Edison) batteries were the first widely used type of storage battery. These original Edison batteries were heavy, slow and very expensive but they did last for a very long time and were totally resistant to deep discharges and being left in partial states of charge. Just the simple fact that the Edison batteries did not ever need to be fully charged continued to make them popular for small off grid power systems long after lead acid batteries had totally taken over the market for all applications.
In the 1960's NASA developed the nickel cadmium batteries as the first modern alternative to lead acid batteries. The significantly higher energy density was the advantage that NASA was after for use in space, but nickel cadmium batteries also had the same advantage of not needing to periodically come to a full state of charge. Dry cell nickel cadmium (NiCad) batteries that came on the market in the 1970's did not work well at all mostly because the chargers supplied with them routinely overcharged the cells causing premature failure. Some flooded nickel cadmium batteries were produced, and worked quite well in the very expensive Peugeot and other electric cars of the 1980's and 1990's. Ultimately though the high cost and environmental problems of the cadmium meant that nickel cadmium batteries were a poor choice for most applications.
The alternative to the expensive and hazardous cadmium was the nickel metal hydride (NiMH) batteries that were first sold in the early 1980's. Nickel metal hydride cells are remarkably similar to the old nickel iron Edison cells, but the use of an iron containing complex ion instead of metallic iron means radically improved performance and much higher capacity. The nickel metal hydride batteries have attained higher energy densities than lead acid or nickel cadmium batteries and also can deliver quite high maximum discharge currents, but they have never been widely available in sizes appropriate for most applications. The small dry cell nickel metal hydride batteries have been very popular for portable appliances, and they can easily hold up to 1000 cycles if they are not overcharged. As with the nickel cadmium dry cells the chargers for the dry cell nickel metal hydride batteries were notorious for overcharging, normally killing the cells within a few dozen cycles.
Aside from chargers for small cells being totally non-functional the big complaint about nickel metal hydride batteries has been that they have not been widely available in larger sizes that would work for off grid power systems, boats or electric vehicles. High 20 or even 30% per month self discharge rates have also plagued both the nickel cadmium and nickel metal hydride batteries, but recently low self discharge rate nickel metal hydride batteries have been available in small sizes. As far as energy density goes nickel metal hydride batteries were for many years thought to be about 50Whr/kg, but recently higher energy density 80, and even 90Whr/kg nickel metal hydride cells have been widely available in small sizes. The technology appears to hold promise, but nickel metal hydride batteries have been overshadowed by the apparently much higher performing lithium ion chemistries. Which is ultimately better for economical electrical storage remains up in the air. At this point lithium ion batteries are cheaper and work much better.