Scheduling for Performance-Decaying Cracking Furnace Operation with Consideration of Inherent Process Upset Reduction

Scheduling for Performance-Decaying Cracking Furnace Operation with Consideration of Inherent Process Upset Reduction^†

Chuanyu Zhao and Qiang Xu*

Dan F. Smith Department of Chemical Engineering

Lamar University, Beaumont, TX 77710, USA

Abstract

 

Thermal cracking furnace system is the critical starting sector of an ethylene plant.  Its multiple furnaces periodically experience hot and cold shutdowns for maintenance due to performance-decaying operational characteristic, which inherently upsets the downstream flowrate of almost the entire plant.  In previous work, most studies are focused on the pyrolysis process simulation and control, or the cracking furnace system scheduling of profit enhancement.  The operational scheduling with consideration of inherent upset reduction is totally ignored.  In this paper, a mixed-integer linear programming (MILP) model is developed to obtain the optimal schedule for any customized scheduling period with two objective functions: 1) maximum average profit of the cracking furnace system per time: 2) minimum inherent process upset from the furnace system.  The idea of inherent process upset reduction is to group shutdown operations of different furnaces together during the manufacturing so that the downstream will have less frequency to suffer upsets from the furnace system. 

The MILP model addresses all the major scheduling issues of a cracking furnace system, such as multiple feeds, multiple furnaces, various product prices and manufacturing costs, hot shutdowns for decoking, cold shutdowns for mechanical maintenance, semi-continuous operation, cracking performance decaying, non-simultaneous shutdowns, and secondary ethane cracking.  With the given customized scheduling period, optimal scheduling solutions for the developed MILP model can be obtained to provide quantitative answers to the following questions: i) what is the maximum average profit of the cracking furnace system within given time period? ii) how to allocate different feeds into different furnace for cracking? iii) how to setup each batch operation and determine its run length? iv) how to determine the best decoking sequence and maintenance location among multiple furnaces? and v) how to manipulate shutdowns from all the furnaces to reduce the inherent upsets to downstream process. 

A case study shows that the significant flowrate upsets from a cracking furnace system can be reduced to two during 150 days scheduling period with the developed methodology, compared to six upsets without the consideration of inherent upset reduction at the cost of affordable profit loss.  It demonstrates the efficacy of the developed methodology in terms of its significant economic potential and operational benefits.  Meanwhile, this work can be applied to dynamic scheduling as well. 

Keywords: Scheduling, MILP, Cracking furnace, Shutdown, Optimization, Upset reduction