부산대학교 도서관

Integration of an energy storage system (ESS) into a large-scale grid-connected photovoltaic (PV) power plant is highly desirable to improve performance of the system and overcome the stochastic nature of PV power generation. Algorithms to size ESS within an integrated PV and ESS ($\mathrm{PV}+\mathrm{ESS}$) power plant, conventionally, require a large number of high-resolution samples of PV generation. These algorithms are computationally expensive. In this paper, a two-day optimization algorithm that utilizes n-step constant power output dispatch every day from the $\mathrm{PV}+\mathrm{ESS}$ power plant is proposed to size the ESS. Additionally, an n-step power dispatch every day through the one-year timeframe based optimization of the size of ESS is also performed. The proposed methods are computationally cheaper and provide acceptable accuracy with respect to the conventional sizing methods. The economics identified from the proposed methods are compared with the typical day-ahead hourly dispatch-based power production from the $\mathrm{PV}+\mathrm{ESS}$ power plant. This comparison has shown similar economic performance showcasing the accuracy of the proposed methods in sizing the ESS. It also shows the potential for operating $\mathrm{PV}+\mathrm{ESS}$ power plant in n-step constant power generation mode.