Development of predictive models for the assessment of the wind turbine foundation's lifespan and structural integrity using ML algorithms

Abstract

Wind turbine foundations are critical components of renewable energy infrastructure, designed to support substantial dynamic loads and withstand environmental conditions throughout their operational life. While wind turbine superstructures typically have an operational life of 20 years, the reusability of the foundation beyond this period remains largely unexplored. This research addresses the ageing and degradation of wind turbine foundations, focusing on developing predictive models to assess their long-term structural integrity and potential for reuse. Utilizing advanced numerical modelling techniques, structural health monitoring (SHM) and material property validation, this study will generate a dataset to determine the ultimate limit state (ULS) capacity of wind turbine foundations at various stages of ageing. This dataset will be used to train machine learning algorithms, resulting in developing predictive models that can inform decision-making on the reusability of ageing foundations for new wind turbines. The project also aims to validate models for concrete ageing and degradation with real-world data from in-situ structures, enabling realistic assessments of foundation performance over time. Additionally, the research will explore innovative foundation designs, retrofitting strategies, and strengthening techniques to extend foundation life, thereby contributing to a more sustainable energy infrastructure. The outcomes will provide valuable insights into foundation reusability, SHM integration, lifespan optimization, and cost-effective design strategies for the renewable energy sector.