MINLP Model for Simultaneous Scheduling and Retrofit of Refinery Preheat Train

• Main Authors: Zulkafli, Nur Izyan, Mahadzir, Shuhaimi, Hashim, Haslenda
• Format: Conference or Workshop Item
• Language: English
• Published: 2013
• Subjects: T Technology (General); QA Mathematics; TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.utem.edu.my/id/eprint/10538/1/PSEAsia2013-80.pdf; http://eprints.utem.edu.my/id/eprint/10538/
• Institution: Universiti Teknikal Malaysia Melaka

There is greater awareness today on the depleting fossil energy resources and the growing problem of atmospheric pollution. Engineers are developing practical techniques to ensure energy processes are designed and operated efficiently. Inefficient heat exchangers lead to higher fuel demand and higher carbon emission. This paper presents mixed-integer nonlinear programming (MINLP) model for simultaneous cleaning and retrofit of crude preheat train (CPT) in oil refinery plant. The formulation of the model is generated and coded in General Algebraic Modeling System (GAMS). The model minimizes the cost of energy and the cost of cleaning. The model takes into account the changes in fouling rates throughout time. There are two cases for this study. The cases are online cleaning (Case 1) and simultaneous online cleaning and retrofit (Case 2). The largest energy saving are found in Case 2. The installation of high efficiency heat exchangers improves furnace inlet temperature (FIT) from 215oC to 227oC. Furthermore, Case 2 results in the highest percentage of cost saving by about 59%. The payback period for investment in high efficiency heat exchangers is 5 months. Thus, Case 2 is the most cost effective option for reductions of energy consumption in Crude Distillation Unit (CDU).