Publication

Abstract : In the future, solar energy is likely to become the preferrable energy source among renewable energy sources to meet out power supply demand. The government of India is planning to increase the use of solar energy for electric power generation to 120GW capacity by 2022 and 450GW capacity by 2030. So the photovoltaic power generation is likely to play an important role when it comes to electricity supply in India. The construction, maintenance, safety, and reliability of the solar power generation sector are all important factors for solar plant operators. A solar plant consists of good-quality PV panels, converter, inverter with connecting wires, and utility meters combined with a monitoring system. Moreover, the predictive maintenance of solar plants is important to enhance the lifetime and efficiency of a plant. Lack of awareness about PV plant maintenance leads to less efficiency. The importance of predictive maintenance includes regular monitoring, performance analysis compared with with previously collected historical data, and prevention of possible failures and energy loss in generation. The advantages of predictive maintenance of solar plants include safety during the entire life span of the plant, reduction in repair and (consequently) idle time, and savings on maintenance and spare parts. In predictive maintenance, regular inspection and remote monitoring systems with sensors are enabled based on recommendations from equipment manufacturers. The main challenges for predictive maintenance of solar plants are identifying immediate changes in performance behavior and remote-control system to control significant parameters such as active power control, reactive power control, frequency control, and voltage control. The recent approaches to predictive maintenance include three methods. First is direct visual periodic inspection of all components, I-V characteristics analysis of the entire PV plant, infrared thermography of the plant and present generation comparison with actual generation capacity of that plant. Second is machine learning and artificial intelligence–based forecasting methods, which are moderately efficient when it comes to detection. Third is smart remote monitoring and control units with wireless sensors networks. Out of these three methods, the wireless sensor method is the most expensive, but it offers the best results in predictive detection. This chapter deals with each component monitoring as well as performance monitoring and various predictive approaches and opportunities for PV plant maintenance to improve the performance of the PV system.