Aging-aware battery dispatch optimization for grid applications

• Main Author: Kumtepeli, Volkan
• Other Authors: Wang Youyi
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering::Electric power::Production, transmission and distribution
• Online Access: https://hdl.handle.net/10356/137747
• Institution: Nanyang Technological University

Day by day the need for autonomy, efficiency, reliability, and sustainability in the power generation, drives researchers to seek alternative energy resources. Therefore, the demand for renewable energy sources (RES) is increasing worldwide, which is followed by the rising contemporary challenges due to the non-dispatchable nature of some of the RESs e.g. solar and wind. In this context, battery storage systems (BESS) appear as a remedy to attenuate the fluctuations caused by the variable power generation. However, despite their rapidly declining prices and improving capabilities, BESSs are yet to be economically feasible in most applications. Thus, their optimized operation plays a major role in reducing their cost. As reported in the literature, there is still a wide gap between BESS and optimization studies. Namely, including state-of-the-art battery aging models within the power system optimization has not been investigated sufficiently. This study presents the construction of an aging-aware energy management system (EMS) for stationary BESSs based on lithium-ion batteries, specifically, lithium-iron-phosphate cells. The EMS incorporates battery degradation costs by using high-fidelity semi-empirical battery models along with a mixed-integer predictive control framework. The proposed method is demonstrated on several grid applications such as energy management in island microgrids and battery dispatch in energy arbitrage markets. Within the island microgrids context, a unique objective function involving various fundamental elements is constructed, and an optimal operating point is pursued using mixed-integer quadratic programming. By leveraging an advanced battery model within the proposed optimization algorithm; diesel generator usage as well as carbon emissions, are minimized. Subsequently, by utilizing in-house experimental results, the cycle-induced battery degradation along with the variable power electronic efficiency is incorporated into the proposed formulation for the optimal battery dispatch problem in energy arbitrage markets. The results highlight that the power electronics efficiency curve has a significant impact on not only the net profit but also the gross profit. Therefore, it should be taken into consideration during the optimization procedure. Lastly, a battery storage system could not be completely represented without adding the temperature effects. Therefore, the used battery model is detailed via removing constant temperature assumption. Hence, the study is finalized by presenting the holistic optimization approach. In summary, this thesis contributes to the literature by investigating the use of advanced lithium-ion battery models in the optimization to pave the way for efficient and effective use of stationary battery storage systems. Hence, it helps to build another bridge between battery and power system optimization researchers.