Discrete-time Scheduling Model of Entire Refinery with Multiscale Operation Time

Abstract

Enterprise-wide optimization is a newly emerging area and has become a major goal in the process industries due to the increasing pressures for remaining competitive in the global marketplace. However, the current researches on refinery scheduling mainly focus on the studies of three sub-problems. In this paper, we present a comprehensive integrated optimization model that includes crude oil scheduling, production unit scheduling, batch gasoline blending, and diesel online blending. It involves two common oil blending methods(batch blending and online blending), and some detailed production characteristics, such as considering influences of different crude oil on the production mode of distillation columns and considering mode transition process of production units. Such a multi-stage chemical process contains various lengths of processing time at different stages. Traditional discrete-time scheduling modeling methods use the greatest common factor of these processing times as discrete-time interval length, which often leads to a huge model size. Based on discrimination of units’ states, we present a modeling approach that can increase the length of the discrete-time interval, so as the size of the model can be effectively reduced and no down-time on units, and the solution time can be decreased ten thousand times compared to the traditional model when achieves a similar objective. Finally, we show the detailed scheduling result and illustrate the effectiveness of the model through three cases.