Heuristic rescheduling for managing abnormal refinery supply chain events

Abstract

Todaya??s many uncertainties require that, for sustainable operations, supply chains be resilient and nimble while reacting to disruptions. We look at a refinery supply chain where disruptions such as crude arrival delay could make the current schedule infeasible; so rescheduling of operations is important. Refinery scheduling for a real world industrial size problem typically requires a significantly large amount of time to generate an optimal schedule. We present a strategy for recovering from disruptions by heuristic rescheduling. The main ideas of our method are breaking down the schedule into operation blocks and preserving the blocks in the schedule as far as possible. The major advantages of our method are greatly reduced computation time, minimum change in handling disruptions as compared to total rescheduling, and multiple proposed schedules. A second method which is model-based is proposed to overcome the shortcomings of the first heuristic-based method. The model-based method is more general, flexible, and complete than the heuristic method. The capabilities of the proposed methods are illustrated with several case studies.