Purpose This study aims to evaluate and enhance maintenance strategies for wind turbines by transitioning from conventional fixed interval preventive maintenance to optimized and predictive maintenance approaches, with the goal of improving reliability and reducing operational costs. Design/methodology/approach The research utilizes real failure data from critical wind turbine components (shaft, gearbox and generator) and models their reliability using Normal, Lognormal and Weibull distributions. Monte Carlo simulations are conducted to analyse system failure behaviour under different maintenance strategies (conventional and optimized preventive maintenance). Additionally, a detailed cost-benefit and return on investment analysis is performed to assess the financial feasibility of predictive maintenance implementation, considering various real-time condition monitoring technologies. Findings Results show that fixed interval preventive maintenance leads to unnecessary maintenance interventions and increased downtime, particularly for components with non-uniform failure patterns. Optimized preventive maintenance, guided by reliability thresholds, significantly reduces redundant maintenance, while predictive maintenance, using vibration monitoring, oil particle detection and thermal analysis, further lowers operational costs, extends component lifespans and minimizes production losses. Return on investment analysis confirms the long-term economic viability of predictive maintenance, despite challenges such as high initial costs and sensor reliability in harsh environments. The transition from regular to optimized preventive maintenance and ultimately to predictive maintenance significantly increases wind turbine reliability and reduces downtime, intervention frequency and maintenance costs. Originality/value This study provides a comprehensive, data-driven framework that bridges reliability modelling with practical maintenance planning in wind energy. It uniquely combines simulation, statistical modelling and financial analysis to demonstrate the transformative potential of predictive maintenance for more sustainable, efficient and cost-effective wind turbine management.
Sadraoui et al. (2026) studied this question.