Wind turbine maintenance is a major cost factor and key determinant of wind farm productivity. Many companies outsource critical maintenance procedures while others perform these tasks in-house, referred to as self-perform maintenance. While expected to reduce time to profit on asset investment, self-perform requires an efficient personnel deployment strategy to implement. In this thesis, a partial solution to the optimization of wind turbine maintenance personnel team assignment is presented. A holistic framework is established, through analysis of historical work orders, for defining metrics that evaluate the performance of technicians. These metrics are further transformed into interpretable proficiency coefficients to be incorporated into an application of the team assignment problem. A case study of a large wind farm owner and operator is presented to illustrate the potential benefits and caveats of the proposed metrics and evaluation strategy. Additionally, the practicality of the data-derived metrics and proficiencies is illustrated. Key improvement strategies in data quality and metric aggregation are detailed, as well as discussion of a potential formulation of the task-to-team assignment problem, to be modeled through a standard maximin approach and solved through an integer programming technique.
Wind turbine maintenance is a major cost factor and key determinant of wind farm productivity. Many companies outsource critical maintenance procedures while others perform these tasks in-house, referred to as self-perform maintenance. While expected to reduce time to profit on asset investment, self-perform requires an efficient personnel deployment strategy to implement. In this thesis, a partial solution to the optimization of wind turbine maintenance personnel team assignment is presented. A holistic framework is established, through analysis of historical work orders, for defining metrics that evaluate the performance of technicians. These metrics are further transformed into interpretable proficiency coefficients to be incorporated into an application of the team assignment problem. A case study of a large wind farm owner and operator is presented to illustrate the potential benefits and caveats of the proposed metrics and evaluation strategy. Additionally, the practicality of the data-derived metrics and proficiencies is illustrated. Key improvement strategies in data quality and metric aggregation are detailed, as well as discussion of a potential formulation of the task-to-team assignment problem, to be modeled through a standard maximin approach and solved through an integer programming technique.
