Wind Farms Maintenance Optimization Using a Pickup and Delivery VRP Algorithm

Operations and maintenance of wind farms in renewable energy production are crucial to guarantee high availability and reduced downtime, saving at the same time the cost of energy produced. While SCADA or NLP-based techniques can be used to address maintenance tasks, efficient management of wind farms can be really achieved by adopting an intelligent scheduling algorithm. In this paper an algorithm that optimizes maintenance intervention routing is presented, taking into account the location of spare parts inventory, geographically dispersed intervention sites, and overall costs of the intervention, considering human resources and fuel consumption. Different scenarios are discussed through a toy example, to better explain the algorithm structure, and a real case of wind farms distributed in Sicily, to validate it. The usefulness of the proposed algorithm is shown also through some Key Performance Indicators selected from UNI EN 15341:2019. The purpose of this work is to show the effectiveness of adopting a VRP algorithm in optimizing the maintenance process of wind farms by investigating real scenarios; in addition, the proposed approach is also efficent therefore feasible for coping with unplanned interventions changes.