From the laboratory: Trends and sensitivities in Hazard Severity Levels of lithium ion cells to common abuse
A prospective buyer or user of a motor vehicle takes for granted that the vehicle has been designed to keep the operator and passengers safe from motor vehicle crashes and other known mishaps such as fire and impact from projectiles. For conventionally-powered vehicles, the passenger safety requirements are well-understood, and design guidelines are based on data from many decades of automotive testing. That long history helps automotive design engineers meet today’s rigorous safety requirements and to find new ways to increase safety levels.
However, the story is a bit different when it comes to hybrid, plug-in electric and electric vehicles (also known as xEVs). Although similar in many ways to conventionally-powered vehicles, xEVs incorporate systems and components that present substantial additional safety hazards, such as high voltage batteries and powertrains, chargers and inverters. These additional hazards complicate the already complex process of designing safe motor vehicles.
The battery of choice for almost all xEVs is based on lithium-ion chemistry. While lithium-ion batteries represent a substantial improvement over other battery chemistries, the total picture of lithium-ion safety performance under severe conditions is still not completely understood. That’s because lithium ion batteries have been used in motor vehicles for less than 20 years, and current design and testing standards are not yet mature enough to ensure safe performance.
One way to remove this uncertainty is to conduct ongoing testing lithium-ion batteries for all possible safety hazards and make improvements where improvements are necessary. TÜV SÜD has been an industry leader in conducting abuse testing of lithium ion batteries, and has conducted more than 1,500 abuse tests on battery cells used to build lithium-ion batteries. The results of this testing has been assembled into a database that allows TÜV SÜD automotive engineers to analyze the testing data to identify trends and to spot specific battery sensitivities.
The data in TÜV SÜD’s lithium ion database represents cells in a wide range of Ah sizes, electrochemistries, geometries, electrolytes and separators, cell build date, and test locations. The data also represents a variety of well-known cell abuse tests such as nail penetration, crush, overcharge, over discharge, short-circuit, thermal stability and thermal extremes. The aggregated date maintains the privacy of individual testing records but also provides an important information resource in our ongoing efforts to improve the safety of lithium-ion batteries.
With such a large database of raw performance data, we can evaluate possible answers to important questions. For example, how does the behavior of battery cells change over time? What specific parameters influence how a cell reacts to abuse? Are the test methods for lithium-ion batteries consistent across all standards? And, how can this information be used to support the development of new and improved standards?
In order to communicate the results in the simplest of term, we rate the results of abuse testing by applying the method developed by the European Council for R&D (EUCAR) and Sandia National Laboratories. This method identifies the results of each test using hazard severity level (HSL) scale from 0 to 7, in which a level of 0 signifies no hazardous effect and a level of 7 signifies an explosion. And HSL of 2 or lower is preferred but a level of 3 or lower usually represents an acceptable level of performance.
As an illustration of the potential value of the TÜV SÜD lithium-ion database, we evaluated the response of single battery cells subjected to nail penetration tests during the most recent three year testing period. The population of cells tested during this period ranged from small to large in size, and were presented in either a hard or soft case prismatic format. All batteries tested were subjected to the same physical restraint conductions observed in battery packs. Figure 1illustrates the results.
Figure 1: Nail penetration: trends over time
As illustrated in the chart, cells subjected to nail penetration tests during the evaluation period demonstrated increasing resistance to abuse, as illustrated by the general increase in the number of tested battery cells resulting in an HSL of 3 or less.
We found a similar trend in evaluating the results from battery cell overcharge testing, one of the most damaging of all abuse tests. Figure 2 once again illustrates a gradual increase in the number of tested battery cells resulting in an HSL of 3 or less. (The red circled data point indicates a misleading favorable result due to insufficient overcharging during testing.)
Figure 2: Three year trend for cell battery overcharging tests
These and other analyses have identified other potentially significant relationships between test parameters and HSL values. At present, our data does not point to a clear relationship between the abuse response of a single cell test and the abuse response of a full battery pack in a motor vehicle crash or fire. Nonetheless, we generally infer that, if the response of a single cell to abuse improves, the abuse response of a full battery pack comprised of those cells will also improve.
It is important to note that these and other battery abuse tests can produce extremely energetic reactions, including explosions, fires and the release of hazardous gases. Therefore, battery abuse testing should only be conducted by trained staff using testing equipment specifically designed for the anticipated hazards and in testing facilities design to accommodate such possible outcomes.
TÜV SÜD continues to analyze the data as new tests are completed to identify other battery trends and sensitivities that may be useful to the battery industry. Watch for further results in future issues of TÜV SÜD’s e-Mobility E-ssentials.