Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > T&M

Model-based approach for safer batteries

Posted: 03 Dec 2007 ?? ?Print Version ?Bookmark and Share

Keywords:battery safety? simulation model approach? battery design?

The pervasive performance and safety problems with batteries can be remedied by adopting a model-based design approach similar to that used for electronic circuits or chemical processes. The battery industry needs to adopt the use of simulation models in order to ensure the quality and safety of their products, and OEMs should lead the way by requiring battery developers to provide simulation models.

This article reviews the current design process, and provides examples of how a model-based approach leads to better-performing and safer batteries.

Designing battery-powered devices today is much more difficult and expensive than necessary. Electrical engineers typically have no idea of how to select batteries. As a result, they rely on battery experts. The situation has led to a process where the device is designed first, and then a battery is sourced to power it.

However, this process all too frequently fails. More seriously, several recent safety incidents have required product recalls.

Need for testing
Disasters usually occur because of inadequate testing or an incomplete understanding of the battery's behavior. Battery developers do not really understand how batteries work as evidenced by their inability to provide physics-based models for battery behavior. Amusingly, battery developers argue on the one hand that batteries are too complicated to model, and then on the other hand, that battery behavior is very simple that no models are needed. The inability to develop physics-based models requires batteries to be empirically characterized. This reliance on empirical characterization inevitably leads to problems.

Battery testing is expensive and time-consuming, and development plans don't usually include enough time or budget for adequate testing. In addition, testing under real world conditions can be difficult. There are an unlimited number of possible use scenarios, so some generalizations must be made to define a finite set of conditions. For example, battery developers typically report voltage vs. discharge capacity at several discharge rates. This information can give a good idea of how a new battery will behave, but all bets are off as the battery ages.

As batteries age, they tend to lose capacity and increase in impedance, and the extent depends on how the battery was used. Several characteristics must be considered (i.e. temperature, state of charge, depth of discharge etc.). The best course is real-time testing of batteries under conditions that approximate the planned use of the battery. This type of testing can easily take a year, but short-cutting this testing can result in a battery that fails prematurely. The empirical approach to abuse testing is even more problematic.

The safety of batteries is determined by testing. Standards organizations have developed abuse tests to identify any potential problem that might result in a battery causing a safety incident. However, as a steady stream of product recalls proves, testing doesn't always work.

The infamous Sony cells that were recalled passed safety tests, but were susceptible to internal short circuits that caused the cells to catch fire. The possibility of internal short circuits occurring was envisioned by existing tests such as crush or nail penetration. However, these tests turned out not to be predictive of an internal short. Without the capability to validate the tests through physics-based modeling, standards agencies were forced to accept the judgment of battery experts that the tests were adequate.

Simulation modeling example
The battery design process needs to be augmented with physics-based models. Developers of battery-powered devices must demand physics-based models from battery developers. By forcing battery developers to demonstrate a mechanistic understanding of their products, the quality and safety of batteries will be dramatically improved.

The following Li-ion cell test examples using simulation modeling focus on comparing chemistries, battery pack design and abuse tolerance of a pack. Before delving into these examples, it's appropriate to discuss some software. Software is required to run physics-based models. The software must be accessible to battery developers and to OEMs, so a third-party provider makes sense. For batteries, the Battery Design Studio software is readily available and is capable of designing a wide range of cells and packs. This software is used in these examples.

Comparing chemistriesNew electrode materials are continually proposed that promise significant improvements in energy density. Battery design software provides a means for evaluating and comparing proposed materials. For example, 18650-sized cells can be designed with different materials, resulting in different energy densities. A study has evaluated how cathodes based on metal phosphates compare to the oxide materials currently used. The availability of standardized software allows the results from such a study to be readily shared.

Pack designOEMs use packs (cells + electronics + packaging + optional heat-transfer means) in their devices, so battery pack behavior is a vital concern. Experimental characterization of pack behavior is complicated by the need to monitor each cell. Modeling can be used to examine how the pack will behave under a wide range of conditions.

For example, one test has modeled the behavior of a battery pack used in a notebook computer and showed that the neighboring electronics affected the temperature distribution, while another test has modeled a more complex pack used by the U.S. Army that showed simulation results were in line with thermal imaging experiments (Figure 1).

Figure 1: Comparison of thermal imaging to simulation results for a battery pack discharged at 10A (at 0.7hrs).

Once confidence has been established in the simulation model, a wide range of use scenarios can be explored. For example, a pulse discharge of the battery pack can be simulated and the maximum cell temperature determined.

Abuse tolerance of a packThe previous example showed that the temperature of a pack could be predicted with reasonable accuracy. This information is of interest in predicting the life of the pack as the hottest cells will tend to fail first. Of even greater concern is the effect of thermal runaway. If one cell in the pack goes into thermal runaway, will the entire pack follow?

Simulation provides a straightforward approach to answering this question. The self-heating rate of a cell can be experimentally measured using accelerating rate calorimetry (ARC). The ARC experiment provides a relationship between the rate of temperature change and the temperature of the cell which can be used to estimate how cells in a pack behave.

A slight improvement in the heat transfer rate could make the pack much more stable against abuse. The approach used to generate Figure 2 is particularly simple, a more rigorous analysis could be used to predict thermal runaway over a wider range of conditions. In this example, as in the previous example, the pack simulations are based on a simulation model for a cell.

These examples demonstrate how simulation can be used to design better performing and safer batteries. OEMs should require their battery suppliers to provide simulation models, and then verify that they are accurate by comparing simulation results to experimental data. If discrepancies between the simulations and experimental data are found, they should be reported to the battery suppliers so that the simulation models can be improved. This iterative process (the scientific method) allows the complexity of battery systems to be conquered and enables real engineering of battery systems.

Figure 2: The pack simulations are based on a simulation model for a cell.

A benefit of using third-party software for this process is that it enables development of standardized analysis procedures. For example, automakers through the U.S. Advanced Battery Consortium have set certain targets for batteries and developed procedures for evaluating how well batteries meet these targets.

One type of hybrid electric vehicle requires that the battery be capable of providing 25kW pulses for 10s over a state of charge range corresponding to 300W/h. The procedure to validate this behavior is somewhat complex, but was programmed into the Battery Design Studio software so that battery developers can quickly estimate how well their products meet the automakers requirements.

The battery industry needs to adopt the use of simulation models to ensure the quality and safety of their products. Taking such a step will put battery development on a path that leads to reduced costs, shorter development cycles, more valuable products and expanded markets.

- Robert Spotnitz
Battery Design LLC

Article Comments - Model-based approach for safer batte...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top