Abstract:
An effective way to reduce energy usage in a refinery is to design efficient heat exchanger networks (HENs) by using process optimization on the Mixed Integer Liner Progeaming (MILP) method. In this work, retrofit designs with/without relocation of HENs are done by GAMS (General Algebraic Modeling System) software. This methodology can generate networks where utility cost, heat exchanger areas and selection of matches are optimized simultaneously. In addition, the simplicity in model assumption of non-isothermal mixing with constraints such as stream splitting and allowed/forbidden matches make the model structure more convenient to use. This MILP model can be successfully applied to the crude refinery, providing both retrofit designs with/without relocation. In a special secnario, relocation topology can be used for further reduction in total cost, which also gives the highest annual cost saving for retrofitting HENs. This research also determines the best HENs for light, intermediate and heavy crude refinery.
