Markov-Based Reliability Analysis of Hybrid PCS: Evaluation of the Effect of Redundancy Design, PV Generation, and Power Demand

Abstract

A hybrid power conversion system (PCS) supplies power to the grid using power sources with different characteristics, such as photovoltaic (PV), energy storage system (ESS), and polymer electrolyte fuel cell (PEMFC). If the output of each power source is connected in parallel to the dc-link and a converter with the same power rating is used, a redundancy effect can be obtained. The operation of the hybrid PCS system is determined by the state of charge (SOC) of the ESS battery, and the operating power is determined by the PV generation and grid demand power or ESS battery charging demand power. In this paper, we design a Markov model that considers the reliability of the hybrid PCS by the operating mode, the operating power according to the PV generation and demand power, the operating power change due to partial failure, and the on-load and standby redundancy methods, and analyze the reliability along with the verification of the operation through simulation to present design guidelines for securing high reliability of the hybrid PCS system. The result of the reliability analysis based on an operating temperature of 50 degrees C is that the standby method's lifespan is 9.8% longer than the on-load method. Also, when PV generation exceeds demand power, the average lifespan is 23.4% longer than below the demand. When demand is 3 kW, the lifespan increases by 31.8% compared to 5 kW, and when demand is 1 kW, the lifespan increases by 12.1% compared to 3 kW.