Extension of Single Particle Model with electrolyte and Temperature (SPMeT) for Real-Time Performance and Safety Monitoring of Battery Energy Storage Systems (BESS) in Grid Service

Abstract

In modern smart grid systems stability and reliability are important criteria due to the inclusion of multiple distributed energy sources such as photovoltaic (PV) and wind power generation. Battery energy storage systems (BESS) are key in enabling this objective of stable grid operation through ancillary grid services such as frequency regulation and energy arbitrage [1] [2]. However, these systems are subject to frequent transient operation under varying environmental conditions leading to significant performance degradation. Accurate prediction and control of battery charge and discharge processes is essential to ensure the safe and reliable operation of these systems. This work focuses on developing a controls-oriented model for Nickel-Manganese-Cobalt oxide (NMC) chemistry lithium-ion batteries based on the existing single-particle model with electrolyte and temperature (SPMeT) from the literature. This work extends the SPMeT model by adding aging mechanisms due to lithium plating on the cell anode during the charging process along with SEI growth. The SEI growth rate was attributed to both capacity and power fade. The lithium plating state tracks the loss in the inventory of cyclable lithium leading to a nonlinear drop in the capacity of the cells. Concurrent with these mechanisms the model tracks the growth of dendrites on the anode as a function of lithium plating. These aging mechanisms enable mitigation of aging for grid-connected cells by active management of charge profile in a model-based controls scheme. The dendrite growth estimate can be used for prognostics and schedule replacement of the cells to prevent battery fires. This model was calibrated using the data from LG M50T cells from automotive applications and later modified to work for grid duty applications. The resultant model is real-time compatible and battery management system (BMS)-friendly and can be used to study the impact of different grid duty cycles on battery life.