Sequential and simultaneous approaches for synthesis of multiperiod heat exchanger network

Abstract

Global energy demand has increased continuously since the last few decades, thus the energy conservation is a critical issue in most chemical processes. Heat exchanger network (HEN) can help improve heat recovery by transferring heat from hot to cold streams. In this work, the synthesis of multiperiod HENs, where the operating conditions are fluctuated, was studied. The sequential and simultaneous techniques based on the stage-wise superstructure model were proposed. The sequential approach consists of three steps. First, a mixed-integer nonlinear programming (MINLP) model was used to generate an initial HEN for a chosen period. Second, the initial HEN was adapted by nonlinear programming (NLP) or MINLP model to generate HENs which were fitted to other period conditions. Lastly, HENs for each period were integrated to obtain the multiperiod HEN design. In the second step, one of three models which have different strategies was used. The most effective model incorporating with a proper starting period was investigated. For simultaneous method, an MINLP model took into account all periods concurrently and was solved in one step. The simultaneous approach showed better performance than the sequential approach; therefore, the simultaneous model was applied further to the case study of crude preheat train in crude distillation unit to assure its performance when dealing with large problem. Furthermore, an initialization strategy was proposed to find an initial feasible solution. It showed that the initialization technique could reduce computational time substantially. Moreover, the solution will be validated by using PRO/II to affirm its feasibility in real process.