A Novel Priority-Slot Based Continuous-Time Formulation for Crude-Oil Scheduling Problems

The optimal scheduling of crude-oil operations in refineries has been studied by various groups during the past decade leading to different mixed integer linear programming (MILP) or mixed integer nonlinear programming (MINLP) formulations. This paper presents a new continuous-time formulation, called single-operation sequencing (SOS) model, which can be used to solve the crude-oil operations problem introduced by Lee et al. 1 . It is different from previous formulations as it requires to postulate the number of priority-slots in which operations take place instead of specifying the number of time intervals or event points to be used in the schedule. This MINLP model is also based on the representation of a crude-oil schedule by a single sequence of transfer operations. It allows breaking symmetries involved in the problem, thus tremendously reducing the computational expenses (all instances can be solved within 2 minutes). A simple two step MILP NLP procedure has been used to solve the non-convex MINLP model leading to an optimality gap lower than 5% in all cases. †Carnegie Mellon University ‡Total France