A new modeling approach to include EAF flexibility in the energy-aware scheduling of steelmaking process

When properly scheduled, steel plants can exploit their flexibility to minimize their energy cost and increase their revenue by participation in demand-response programs aiming at increasing the grid stability. Electric arc furnaces (EAFs), as the most energy-consuming component in such plants, can play an important enabling role in this regard. While advanced mathematical tools such as Mixed Integer Linear Programming have been proven efficient in handling the complexity of scheduling problems, the integration of energy-related constraints has always been a major challenge. In this work, we are going to propose an extension to the state-of-the-art optimization models, such that the flexibility of the EAFs can be further exploited. To this end, first we use the concept of virtual machines as a natural extension to a state-of-art model in literature. Then a novel scheduling model is formulated to include multi-mode operations for the EAFs. Numeric experiments are carried out to investigate the tradeoff between optimality and computational performances and highlight the effectiveness of the multi-mode approach for industrial-scale problems.