• A unified continuous-time framework is developed for enterprise-wide scheduling and planning optimization • Continuous-time representation eliminates the time-scale inconsistencies between planning and scheduling • Three representative continuous-time formulations are systematically evaluated in terms of optimality and computational efficiency • The unit-slot formulation achieves superior efficiency and solution quality Inconsistent continuous-time and discrete-time representations have been a big challenge in achieving the enterprise-wide optimum for integrated crude scheduling and refinery planning problems. This paper proposes a unified continuous-time representation for solving the integrated enterprise-wide optimization problem. Instead of implementing continuous-time slots for the crude scheduling subproblem and discrete-time slots for the refinery planning subproblem in a hybrid way, continuous-time slots are applied to both subproblems and synchronized via the crude distillation units. Such a proposed modeling framework provides a more flexible and consistent time formulation to the integrated problem. Furthermore, three typical continuous-time representations for crude scheduling, namely, unit-slot time representation, event-based time representation, and multi-operation sequence time representation, are investigated with cases adapted from practical complexes. Compared to the traditional integrated problem based on a hybrid continuous-time and discrete-time representation, the proposed unified model framework reveals a better optimization performance. Higher-quality solutions can be obtained for both crude scheduling and planning operations. Moreover, among the three time representations, the model based on the unit-slot time outperforms the other two models. This work provides a promising alternative modeling strategy for achieving enterprise-wide optimality in the petrochemical industry.
Wang et al. (Sun,) studied this question.