A Stochastic Optimization Framework for Condition-Based Maintenance, Crew Management, and Spare Parts Logistics in Solar Energy Systems

Abstract

This study presents a stochastic optimization framework for condition-based maintenance scheduling in solar energy systems, considering production, spare parts logistics, and crew management. The proposed framework leverages a detailed model to optimize hourly operations over a 30-day timeframe with a rolling-horizon fashion, integrating maintenance signals derived from the remaining lifetime distributions of each photovoltaic (PV) unit across various locations. The objective of the model is to minimize the expected cost of preventive and corrective maintenance activities, inventory holding and order, crew routing, and penalty for deviation in the crew routing plans while maximizing expected production revenue. Expected revenue is calculated through hourly and daily electricity prices and production scenarios generated from real-world data for each location, while the expected costs are obtained using uncertain maintenance duration. This study offers a comprehensive approach to asset management through efficient crew management and spare parts logistics, leveraging opportunistic maintenance and optimal crew routing between different asset locations, while balancing costs and reliability. The proposed approach provides significant advantages over the benchmark policies regarding efficiency and reliability.