Multi-period Planning of Refinery Crude Oil Scheduling Problem with Fluctuating Demands

In this paper a multi-period planning and scheduling problem under demand correlations is addressed. A stochastic optimization model using restricted time structure is formulated. This model involves the minimization of the operation cost subject to constraints for the satisfaction of crude mix demands with a prespecified level of probability. A chance constrained programming that converts the stochastic programming to a deterministic one is developed for solving the problem. The deterministic problem is then solved by a proposed approximation algorithm iteratively. Finally, case studies are effectively solved by the proposed approaches. The results shows that the cost can be effectively reduced compared with the operations without considering the demand correlation.