Planning and Management of Energy Storage in Microgrids according to Complex Network Approach

This PhD dissertation deals with planning and management of Energy Storage System (ESS) in power systems. Chapter 1 is devoted to the State of the Art of ESS technologies, discussing them according to their structure and physical and chemical proprieties. For each class, the most relevant ESSs are presented and briefly portrayed, featuring their main advantages and drawbacks. In this way, a general comparison is performed with regards of different characterization of ESSs. Chapter 2 refers to planning and management of ESSs in a power system and presents the genetic algorithm-based multi-period optimal power flow (GA-MPOPF) a management method that includes degradation costs in the economic optimization of ESS. The GA-MPOPF includes complex aging function taking into account cycle and calendar degradation costs. This is still an open issue, mainly due to the difficult estimation of cycle aging in presence of high penetration of intermittent Renewable Energy Sources (RESs). The proposed method has been exploited in two different microgrid applications based on real load and generation data: (1) planning of the optimal position of a Li-ion battery ESS (BESS) in the standard 69 IEEE bus network with high RES penetration; (2) the economic management of BESS including cycle and calendar costs. Results demonstrate that GA-MPOPF can optimize the BESS use for one month, notwithstanding the complex operative costs functions, guaranteeing in the meantime excellent convergence properties. Chapter 3 deals with Real-Time planning and management of ESSs in the power systems. The presented GA-MPOPF has been extended to the case of Real-Time aging cost management by including a prediction algorithm based on system identification methods.Two main goals are considered: first, to have a correct BESS size within planning approach; real-time management of the BESS State of Charge (SoC) by taking into account the its degradation costs minimizing the total costs for the whole microgrid. Results show that the accuracy of cost optimization in real time is comparable with the ideal case of a perfect knowledge of the future (i.e. using for Real-Time optimization the actual data). Results also confirm that the method is able to optimize complex cost functions highlighting that a careful sizing of BESS is need to avoid economic losses due to BESS aging costs. Chapter 4 introduces a criterion for the optimal placement of active and reactive power compensator (i.e. ESS) based on complex networks centrality metrics aiming at voltage regulation. The chapter shows the relation between centrality measures and voltage fluctuations in power networks with high penetration of RESs and ESSs. In fact, the correlation between network node centrality (namely Eigenvector, Closeness, Pagerank, Betweenness) and voltage fluctuations is statistically significant implying that the topological characteristics of the power networks are enough to find the optimal positioning of active and reactive power compensators. The results demonstrate that eigenvector centrality shows a statistically significant exponential correlation as the voltage stability increases. This finding provides a quick and easy way to position reactive power compensators in complex networks without the need to compute the traditional Optimal Power Flow (OPF).