Robust condition-based generation maintenance: Balancing operations and start/stop cycling to control asset degradation rates

The integration of renewable energy, distributed generation, and electric vehicle charging into modern power grids has created a highly variable operational environment. Current operations and maintenance models lack the flexibility to accommodate this increased variability and its effects on degradation, leading to more frequent start/stop cycles that significantly impact asset lifespans. In this paper, we propose a robust optimization framework designed to optimize generation maintenance schedules and unit commitment decisions in power systems. Our approach explicitly models the impact of start/stop cycling on remaining useful lifetime distributions, providing accurate failure risk assessments. By incorporating sensor-informed and decision-dependent degradation models within an operations and maintenance optimization model, our framework effectively balances lifetime utilization, failure risks, and operational efficiency. We validate the effectiveness of our framework through computational experiments using real-world degradation signals, demonstrating its advantages over the benchmark models.