The goal of this research was to determine how temperature gradients in series-connected lithium-ion battery packs affect the performance and cycle-life degradation of each of the cells in the battery pack as well as the pack as a whole. This degradation was compared to cells cycled individually and cells subjected only to calendar aging. This allowed attribution of degradation rates to specific usage conditions. Cell performance and capacity degradation was measured by utilizing standard capacity and resistance checks, neutron radiography, and other battery performance metrics. It was found that cells subjected to non-uniform temperatures within the pack degraded faster, and battery pack capacity was reduced relative to an ideal battery pack. It was also determined that the hottest cell in a series connected pack with a temperature gradient will degrade faster than an equivalent cell cycled individually at the same temperature.

Despite the presence of large battery packs in modern electric vehicles, little work has been done to verify pack performance and degradation models, especially when paired with non-uniform pack temperature. This was due to the interwoven degradation stressors and mechanisms at play that need to be separately investigated to gain a full understanding of the combined effect of series packs and temperature gradients. In this work, the effects of calendar aging, cycling aging, temperature, and temperature gradients were all measured and compared. The causes of increased degradation measured on the packs with an applied temperature gradient were determined to be the superimposed effects of varying depths of discharge and non-uniform current distributions within the cells. It was also found that cell degradation did not scale linearly with applied temperature gradient. Rather, the relative rate of aging was found to follow a second order polynomial in temperature and in temperature difference. This meant that depending on the pack average temperature, the hottest or the coldest cell can degrade the most. This behavior was quantified to allow for use in optimizing the design of battery thermal management systems, and an example analysis of thermal management performance was performed on a battery pack from a Nissan Leaf.