Energy-aware flexible job shop scheduling under time-of-use pricing with renewable, battery storage and preventive maintenance

Improving the energy cost-efficiency of manufacturing systems has become increasingly important with the adoption of time-of-use (TOU) electricity tariffs, on-site photovoltaic (PV) generation, and battery energy storage systems (BESS). In flexible production environments, however, the effect of these energy resources depends strongly on the interaction between routing flexibility, operation sequencing, maintenance constraints, and the temporal structure of electricity prices. As a result, the benefits of renewable integration, storage, and scheduling decisions may vary significantly across system configurations and operating conditions. To address this issue, this paper proposes an integrated optimisation framework for the Energy-Aware Flexible Job Shop Scheduling Problem (EAFJSP) under TOU pricing, preventive maintenance, PV generation, and battery storage. The objective is to jointly minimise makespan and grid electricity cost within a unified decision framework. To this end, an exact Mixed-Integer Linear Programming (MILP) model is developed for benchmark instances, together with a scalable Genetic Algorithm (GA) combined with a post-optimisation Energy-Aware Scheduling (EAS) procedure for larger instances. Computational experiments on 13 benchmark instances are designed to quantify the individual and combined effects of PV integration, battery storage, and scheduling-based load shifting. The results show that PV integration can reduce grid electricity cost by about 20-60% without increasing makespan, while the proposed EAS procedure provides additional savings of 1-22% on the tested benchmark set, depending on instance flexibility and operating profile. Sensitivity and energy-flow analyses further identify three operating regimes, clarifying when battery expansion provides greater benefit than scheduling refinement. Overall, the study shows that production efficiency and energy cost reduction can be improved jointly through a unified and operationally interpretable scheduling framework.