Short-term planning model for petroleum refinery production using model predictive control

Abstract

A model predictive control (MPC) strategy is developed to determine the optimal solution of the short-term refinery production planning problem. The main objective of the proposed algorithm is to maximize the total profit and to minimize the costs regarding the refinery process over a planning horizon. The refinery planning problem is solved in discrete time over a pre-determined prediction horizon and this prediction horizon is shifted by one time interval at each subsequent step where the optimization is repeated. To demonstrate the performance of this receding horizon strategy, two literature examples are introduced where the refinery process comprises the oil fields, crude oil vessels, the storage, charge and production tanks, as well as the crude distillation units (CDU). The performance of the strategy for different moving horizon lengths is presented and discussed. © 2009 IFAC.