Fault Mitigation and Cell Balancing of High Power Lithium Ion Battery Packs 2010-01-1766

Lithium-ion (LI) batteries are rapidly becoming a viable choice for military and civil electric vehicles (EV), hybrid electric vehicles (HEVs), unmanned systems, and other applications, mainly because they contain higher energy density, provide higher cycle life, offer better resistance to memory effects, and weigh less than other potential technologies. These same benefits have also led to widespread integration of lithium-ion products into the portable electronics markets. However, lithium-ion batteries carry their own disadvantages, including degradation at deep discharge, capacity loss at high temperatures, and susceptibility to catastrophic failure from venting (especially during charging), shorting, etc. that can have dire consequences on the platform. Another concern with EV/HEV applications is that many cells (packaged as battery packs/modules) are needed to provide sufficient power. This situation leads to thermal and electrical cell imbalances, which significantly reduce the performance of the system. If LI batteries are to be effectively fielded for high power applications (i.e. 10+ kWh), technologies are needed that can be used to actively mitigate the risk of catastrophic failure and ensure proper balancing across the pack.

In order to address this situation, the authors are developing an advanced battery management system for LI battery packs that ensures adequate, safe, and reliable operation. This paper presents the results of experimental and analytical work that is being performed by the authors for both military and commercial applications. Experimental results include discharge/charge cycle tests that were conducted while collecting common measurement signals (i.e. voltage, current) and battery impedance. Advanced electrochemical models have also been developed that aim to capture physical phenomena that cause capacity degradation in cells. The work includes novel methods for impedance measurement, system modeling, SOC/SOH assessment, and advanced prognostic algorithms for remaining life and charge assessment.