Condition-based maintenance policies for Scrap-Based steel production lines

Abstract

Steel production is the backbone of modern infrastructure and technological advancements but remains one of the largest industrial contributors to global greenhouse gas emissions and energy consumption. Recycling steel through scrap offers substantial environmental and economic benefits, including up to an 80% reduction in carbon emissions, significant energy savings, and the promotion of circular economy principles. However, transitioning to scrap-based steel production faces challenges, particularly in meeting rigorous quality standards for high-grade steel and optimizing recycling efficiency. A key obstacle lies in shredder machines, which are critical to scrap processing but prone to maintenance complexities such as component degradation, imbalance issues, and interdependencies between system components and production-line workstations. These challenges adversely affect recycling efficiency, operational costs, final product quality, energy consumption, and industry competitiveness. This thesis addresses these critical issues by proposing two novel condition-based maintenance (CBM) models for shredder systems in scrap-based steel production. The first model integrates multi-level inspections within a hybrid maintenance policy, combining condition-based and age based strategies to address imbalance failures, progressive wear, and external shock occurrences. The second model employs real-time condition monitoring and reinforcement learning (RL) to optimize maintenance decisions in dynamic and uncertain production environments. Both models capture interdependencies among shredder components, enabling enhanced reliability assessments and optimized maintenance planning. Case studies and benchmarking against conventional strategies demonstrate that the proposed CBM models reduce operational costs while improving system reliability and production efficiency. Sensitivity analyses further demonstrate their adaptability to diverse operational scenarios, offering valuable insights for real-world implementation. By advancing maintenance optimization in scrap-based production lines, this research contributes to sustainable steel production, empowering the industry to produce high-quality steel with increased reliance on recycled materials while also aligning with global sustainability goals.