State of Health Estimation Method Design for Energy Storage System of Lithium Ion Battery and Comparative Study : From Cell to Demonstration Sites

State of health Estimation method design for Energy Storage System of Lithium ion battery and comparative study : from cell to demonstration sites

Keonhee Park, Ph.D. candidate
Seoul National University

Recently, Increasing demands of renewable energy and necessity of stable electricity supply are accelerating ESS(Energy storage system) market. ESS, which can stores the power and supplies electricity when it is needed, is possible to increase efficiency of electricity-usage and still has converted from production-consumption paradigm of electricity system to production-storage-consumption paradigm of that so far. ESS is comprehensive system including EMS(Energy Management system), BMS(Battery management system), PCS(Power conversion system). BMS has important role for sustainable ESS operation. Main state variables of BMS are SOC(State of charge) and SOH(State of health). SOC means current residual capacity based on present nominal capacity. And SOH is a prominent indicator to determine degradation of battery. Performance of repeated cycled battery decreased and it should be replaced at an appropriate time for usability. Imprecise SOH can affected SOC accuracy which is important variable for on-line monitoring. However, SOH accurate estimation is challenging due to few obtainable output variables for the battery system and the lack of understanding for electrochemical phenomena in cell. On-line SOH estimation method is needed since a lot of ESS application sites are isolated. This study proposes the modified online SOH estimation method, and compares estimation results of related SOH estimation methods. The results are validated with experimental cell data and demonstration operation data. This method uses the equivalent circuit model parameters and recursive least square method. The cell experimental data is given by Korea testing certification(KTC) and demonstration data is obtained by 2 sites, which are interconnected with PV power system. The one is installed in micro-grid system in the island and the other is integrated with plant site. From the test results show maximum 2% accuracy in cell data and 5% accuracy in demonstration data. The accuracy of proposed method is dependent on input data quality highly. Also we can estimate the life time of battery system and implement easily without complex change of BMS algorithm.