Analysis of Battery Energy Storage System Sizing in Isolated PV Systems Considering a Novel Methodology and Panel Manufacturers Recommended Methodology

This paper presents the analysis of two BESS sizing methodology in solar PV systems, a proposed methodology considering predicted hourly solar radiation data and the methodology recommended by panel manufacturers. In the proposed method, Solar radiation behavior is predicted by studying and processing historical hourly solar radiation data of a location in Brazil using Box-Jenkins method and autoregressive (AR) and time series models are used to generate hourly synthetic series. The generated series combined with hourly load demand and battery storage capacity are used in simulating a PV system and the BESS is sized considering energy deficit and supply interruption outcomes. Comparison is made between the result of the proposed methodology and that of panel manufacturers' methodology considering two case studies. Results of the analysis showed that the proposed methodology is more adequate for BESS sizing. Probability analysis is also performed using multiple radiation scenarios of synthetic solar radiation data.