Development of a system level post prognostics reasoner for FRP turbine blades

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2024
Thesis identifier
  • T16939
Person Identifier (Local)
  • 202063179
Qualification Level
Qualification Name
Department, School or Faculty
  • The surge in offshore wind energy amplifies the urgency to optimise O&M lifecycle costs, a pivotal endeavor for bolstering affordability. These costs are anticipated to constitute a significant portion of overall lifecycle expenditures. A crucial strategy in achieving these cost reductions involves transitioning from traditional maintenance models, such as calendar-based repairs, to more sophisticated approaches like Predictive Maintenance. This paradigm shift poses a formidable challenge, as uncertainties related to damage propagation, weather dynamics, and maintenance planning exert considerable pressure on O&M practitioners. The objective of this thesis is to delineate steps illustrating the design of an autonomous decision-making system for wind turbine blades. The initial phase involves identifying the most consequential failure modes through a comprehensive Failure Modes, Effects and Criticality Analysis (FMECA). Subsequently, a degradation function is proposed for a primary failure mode, namely leading edge erosion, furnishing the groundwork for approaching the O&M optimisation challenge. In the progression toward an autonomous system, an essential tool is introduced to facilitate the selection of baseline calendar-based maintenance strategies for leading edge erosion at the wind farm level. This tool serves as a precursor to the ultimate design of a RL-based autonomous decision-making agent, incorporating prognostics information specifically for leading edge erosion. The obtained results showcase the efficacy of the proposed agent, demonstrating a noteworthy reduction in expected costs ranging from 12% to 21% when compared to condition-based maintenance. Furthermore, the agent contributes to a diminished risk of blade failure, highlighting the promising impact of autonomous decision-making in the realm of wind turbine O&M.
Advisor / supervisor
  • Kolios, Athanasios
  • Brennan, Feargal
Resource Type