Maintenance scheduling optimisation for sustainable steel ladle fleet operations

Abstract

An adequately managed steel ladle fleet is crucial for cost-efficient and sustainable steelmaking. While short-term dispatching decisions have been widely studied, the sustainability of long-term decisions, especially ladle maintenance optimisation, remains unexplored. To address this gap, we present a novel mixed-integer quadratically constrained programming model that captures the complete ladle life cycle, integrating production and refractory degradation constraints to minimise the CO₂ emissions. Exact solvers can solve it efficiently while incorporating operational rules to mitigate end-of-horizon effects. Using historical data from an integrated steel plant, we analyze the refractory degradation patterns and optimise ladle deployment across three production scenarios. Our results highlight the importance of synchronising ladle relining with low-demand periods in the face of production variability. However, this policy introduces challenges in managing personnel and maintenance resources, as multiple relinings may be scheduled simultaneously. Furthermore, direct emissions are reduced by up to 39% when compared to a typical rule-based scheduling approach. These findings provide valuable insights into how an inefficient maintenance policy affects economic and environmental performance. In conclusion, this study contributes to a more holistic understanding of sustainable steel ladle fleet operations, offering actionable strategies for balancing production efficiency, resource consumption, and emissions reduction in steelmaking.