Investigating the Aging Effect of Current Ripple on Lithium-Ion Cells

Within the field of electrochemistry, it is generally accepted that the rate of most electrochemical reactions do not exceed the 1 ms range. In other words, equilibrium perturbations with a frequency of more than ~1 kHz will not affect the chemical composition of a given system. For example, in the case of a lithium-ion cell a high frequency change in applied voltage would be expected to charge and discharge the electrochemical double layer without causing intercalation or deintercalation of lithium ions at either electrode.

However, batteries are applied devices which can be found in virtually every field of technology imaginable. Experts from other fields do not view batteries from the same perspective as the electrochemist, and often ascribe to them preconceived properties which may not be accurate. As an example, the drive train in an electric vehicle includes a large capacitor, called a DC-link capacitor, between the switched mode power converter and the battery. The purpose of this capacitor is in part to shield the battery from residual AC harmonics, the frequency of which typically exceeds 1 kHz. If this shielding is unnecessary, the DC-link capacitor could be made significantly smaller which would simplify the entire drive train.

In order to determine whether or not the current design of the DC-link capacitor is justified, we have built an experimental setup where twelve prismatic lithium-ion cells with a capacity of 25 Ah each will be exposed to a high frequency current ripple during cycling. The twelve cells will be divided into groups of three, where each of the four groups will be exposed to a different type of current ripple. All cells will be cycled with a rate of 1.6C (40 A), and the four groups will be exposed to ripple with a frequency of 100 Hz, 1 kHz, 10 kHz, and no ripple, respectively.

Experiments will begin in May 2015 and run for at least one year, or until the cells under test reach their end-of-life, defined as having degraded to 70% of the original capacity or being unable to sustain a 5C power pulse for 10 seconds. The experiment will periodically be halted to perform characterization tests on the cells. In parallel with the experiments, a mathematical model describing the time dependent behavior of lithium-ion cells when exposed to high frequency perturbations will be created. The model will be experimentally verified using the results from the experiments.

This project will show if the aging properties lithium-ion battery cell aging is negatively effected by high frequency current ripple, of the type commonly found in electrical vehicle drive trains. We hypothesize that the effect is lesser than is currently assumed when designing such systems. If so, it would the DC-link capacitor could be made smaller because it does not need to protect the battery from residual ripple from the power converter. In addition, it may be possible to synergistically use the battery's double layer capacitance to further reduce the size of the DC-link capacitor. A mathematical model which could be used to make such design decisions will be developed.