The BESS was found to be operational 90% of the time and it stored a cumulative 1.5 GWh of energy, which represented more than 5000 equivalent full cycles on the cells. From this BESS usage data, an initial estimate of BESS degradation was provided and a representative duty cycle was developed. The analysis of the maintenance cycles indicated that these 5000 equivalent cycles induced an estimated 5-10% degradation of the single cells. The battery duty cycle was characterized based on 5 parameters: pulses duration, pulses intensity (current), state of charge (SOC) swing range, SOC event ramp rate, and temperature. The average usage consisted of several 9-second C/2 charge and discharge pulses organized to generate 5% SOC swings with a 0.75% SOC/min ramp rate at 35°C. However, extreme values such as currents up to 4C, swings of 100% SOC, and temperatures above 50°C were also recorded.
Laboratory testing and analysis, in conjunction with a more thorough SOH estimation protocol, resulted in a detailed description of degradation that improved the predictions of the remaining useful battery life. Based on the BESS representative usage profile, cycle-aging and calendar-aging experiments were designed to test the degradation of the associated Li-ion cells in a controlled fashion. It was proven that the cell temperature history had the strongest impact on battery degradation followed by the C-rate and the state of charge. Interestingly, the impact of SOC, both on the cycle-aging and the calendar-aging experiments, was revealed to be counterintuitive. During cycle aging, small SOC swings were more detrimental than larger ones. During calendar aging, batteries lost capacity faster at low SOC than at high SOC. The associated degradation mechanisms and their path dependency were analyzed using incremental capacity analysis.