A battery of tests…
13 Jun 2013 by Evoluted New Media
Recent high profile problems with the Dreamliner have highlighted the importance of getting battery technology right – but could Boeing have avoided the issues with the 787’s lithium batteries with appropriate laboratory testing?
Most modern high energy batteries, not only Li-ion, contain highly reactive and potentially explosive chemicals as exemplified by the recent Boeing Dreamliner incident. Testing techniques exist that can predict these problems yet are rarely used and not widely understood. The most important of the tests aim to define the safe working limits: safe temperature, maximum discharge current and maximum safe voltage. Yet some of these parameters are often missing on battery data sheets and the even when they are quoted, there is little supporting evidence.
These important working parameters are often mixed up with the whole range of so-called “abuse” tests that are specified for Li-ion batteries. In reality these particular “abuse” tests are special, both because they are much more fundamental to the use of batteries and also because the data can only be obtained on special equipment, called an adiabatic calorimeter.
The United States’ FAA has listed 132 previous aircraft incidents between 1991 and 2012 that involved “smoke, fire, extreme heat or explosion” in which battery powered devices were implicated and 62 of these incidents involved Li-Ion batteries. The result of the Boeing incident was mostly confined to the battery enclosure and though expensive for the companies involved, it did not lead to any injuries. This will not always be the case and especially when larger numbers of more powerful cells are used, for example in EVs, the situation can be far more serious. Commercial pressures such as the need to drive EVs more quickly, over greater distances and charge the batteries in minutes rather than hours will increase the importance of battery testing.
The importance of testing in an adiabatic calorimeter lies in the fact that it represents a reasonable but worst case situation. When the same test is done, for example to examine limits of overcharging, without such device, the hazard is understated and might even imply the absence of any risk. The other important aspect of custom designed Battery Testing Calorimeters (BTC) is that large batteries (more 50cm x 50cm) and even packs can be tested as they would normally be used and the intensity of the fire or explosion, the amount of toxic and hot gases generated and the speed at which the disaster develops are there to see.
Essentially, these calorimeters allow the heat generated by the mal-function of a battery to be retained within the battery so that its temperature rises in proportion to the heat liberated and thus enables the consequences of malfunction to be realistically and unambiguously measured. In extreme cases, the temperature can continue to rise more and more quickly (often called a thermal runaway) leading to the generation of vast amounts of toxic chemical gases and the battery can physically disintegrate and possibly catch fire too. This is therefore a realistic simulation of an accident, using real battery samples, but performed in a laboratory and possibly photographed.
The most basic adiabatic calorimetry test determines the battery temperature at which problems of thermal runaway start and hence defines the maximum safe temperature. This involves use of a heat-wait-search (HWS) procedure that starts by heating the sample in small steps (see figure 1) and at the end of each step, the system “waits” to see if the battery is generating heat that can be measured by a temperature rise (so called “search” step).
In this experiment a pouch type Li-ion battery was being tested and the “search” procedure starts at around 35oC and since no heat generation is detected (temperature remains constant) the battery is heated again. This stepwise heating followed by a wait period before search, is repeated until self-heating within the battery can be detected (at around 120oC); this is essentially the maximum safe temperature.
Doing the same “abuse” test without an adiabatic calorimeter would have three main drawbacks:
- The correct maximum safe temperature would not be determined. More likely a temperature higher than the safe figure would be obtained (ie the battery would be perceived to be less hazardous).
- The consequences of the thermal runaway would be understated in terms of severity and speed of incident. For example how hot the battery gets, the amount of fumes generated, the damage to the battery itself and of course how long it takes to produce these conditions would be less severe in a non-calorimetry test.
- A custom-built calorimeter provides a safe environment for operators to carry out the tests. The absence of such provisions can present a serious hazard to the operators.
While the thermal stability of normal (undamaged) batteries is of huge interest it is also important to determine how the results change under abnormal conditions or when a battery is charged and discharged at too fast a rate. There are huge commercial pressures to speed up charging/discharging; in the context of EVs, charging is the equivalent of filling a tank with fuel and the discharge rate determines the car speed. If the battery is connected to a cycler while placed in the BTC changes in the battery temperature during charging/discharging cycles can be measured. The cycler can be programmed to repeat this operation for many days and thereby also provide information about the longer term stability of the battery.
A Li-ion polymer battery, used to fly model airplanes, was subjected to higher charging and discharging currents than recommended, while inside the BTC. The temperature rose during discharging and fell during charging, but overall there is was a continued rise in temperature and after only a few cycles the battery went into thermal runaway – the temperature being around 110oC when it did so.
Clearly, at this discharge current, the battery would most definitely need to be cooled to prevent runaway, though the cooling duty is not known from this test.
The overcharing (voltage) test is similar to the over-discharge (current) discussed above except that now the current is at a normal (and safe) value but the battery is charged to higher voltages until it starts to thermally degrade.
Battery technology to tackle current demands, let alone higher energy applications in the future, already has the potential to cause violent fire and explosion incidents. Fully developed testing methods, widely used in the chemical industry for more than 30 years, exist which can reveal the limits of use and demonstrate the hazard. However, at present, far too little approriate testing is being done and as a consequence, accidents will continue to occur. The expensive problems with the Dreamliner could almost certainly have been avoided with a better understanding of the risk through use of adiabatic calorimetery testing with a device such as the BTC.
Author: Jasbir Singh, Founder and Managing Director, HEL Ltd
Contact: www.helgroup.com
