Intrinsic Degradation Variability in Commercial Lithium-Ion Batteries

Lithium-ion batteries have first been introduced to market for consumer electronics applications where safety and reliability requirements are less stringent than in industrial applications. 18650-format cells in particular have encountered great success in the laptop industry and the manufacturing processes have matured to the point where 18650 cells found their way into new markets such as personal transportation or energy storage systems. Because 18650 cells are a commodity, they face typical manufacturing process variability which can be an issue with these larger scale battery systems. Indeed, where a laptop battery consists of 2 to 4 cells in series, the battery found in a passenger vehicle such as the Tesla Model S can consist of a series cells count in excess of 100. When taking into account the need for long-term reliable operation of the battery pack and the cost burden associated with premature battery pack failure, it seems critical to ensure long-term performance of the battery pack. Unfortunately, a battery pack is usually as weak as its weakest cells.

Battery pack’s early-life performance and reliability are usually attained by sorting and matching cells used in a battery pack. This mitigation technique has its limits though as it cannot account for the long-term degradation spread of the individual cells composing the battery pack. This degradation spread within a battery pack can have various origins. One of them can be the temperature gradient resulting from the spatial distribution of cells within the pack and the inhomogeneous cooling architecture. Another one can be the intrinsic variations between individual cells toward degradation sensitivity. These intrinsic variations have been documented in a few studies [1-2] but authors did not offer an elucidation of the underlying mechanism. In this presentation, we explore the nature and quantify the extent of the degradation discrepancies within a batch of commercial lithium-ion cells.

Results will be presented that distinguish performance mismatch down-the-road between kinetic origins (internal resistance build-up, degraded rate capability) and thermodynamic origins (loss of lithium inventory, loss of active material). In a limited previous study [3], we established a qualitative correlation between thermodynamic changes and kinetic properties of cycle-aged cells. In this more comprehensive study, we cycle-aged a larger number of single cells for more than 1,000 cycles to derive statistical correlations between these dual aspects of battery degradation.

Figure 1: Evolution of 15 cells capacities.
Returned capacity spread became very significant with cycling

References:

[1] Baumhöfer, T.; Brühl, M.; Rothgang, S. & Sauer, D. U., Journal of Power Sources , 2014, 247, p. 332
[2] Eom, S.W.; Kim, M.K.; Kim, I.J.; Moon, S.I.; Sun, Y.K. & Kim, H.S., Journal of Power Sources , 2007, 174, p. 954
[3] Devie, A.; Dubarry, M. & Liaw, B.Y., 224th ECS Meeting, Orlando, 2014