Multi-Periods Heat Exchange Network Designs Based On Sequential Synthesis

Heat exchanger networks are widespread in the energy recovery system of process industry, its rationality and efficiency of the designs have a close relationship with the overall performance of the system. With the different heat transfer needs in different periods, in order to improve the efficiency of heat exchange, the heat exchange network structure should be changed. Heat exchange network synthesis problems, i.e., in various allowed conditions, making full use of the energy with the process streams and reducing the utility consumption. At the same time, by further optimizing and adjusting the heat exchange network, the capital cost is saved, and then the minimum total cost of the network is guaranteed. For multi-periods heat exchange network synthesis problem, this paper had carried out research works, the main innovations include:combined the single-period heat exchange network design based on sequential synthesis method with the heat exchanger timesharing design, the minimum total network investment cost was obtained; pipeline optimization model was proposed to reduce the pipeline connections cost between the heat exchangers, optimized the network structure.Firstly, in order to solve large-scale heat exchange network design problems, with the advantages of the pinch technology and the mathematical programming method, sequential synthesis method was used to produce the optimization design for each period of operation in this paper. In turn, LP, MILP and NLP mathematical programming model was established, in GAMS software environment, by choosing appropriate solver for model calculated, the minimum operation cost, minimum number of heat exchange match and minimum capital cost could be respectively obtained in each period. What’s more, the solution time for the large-scale model could meet the actual application. For the same heat transfer match, because of the area diversity in different periods, the maximum area method was chosen to determine the heat exchanger area needed for the whole network, so as to meet the heat transfer demand of the different periods.Secondly, faced with the heat exchanger idled phenomenon after integration, this paper used three strategies to realize the timesharing design for the heat exchangers in all periods. The results showed that for different sizes of heat exchange networks, selected the appropriate share strategy could reduce the capital cost to a certain extent, it could also increase the flexible operation of the network. To overcome the shortage of the above strategies, the third strategy of the mathematical model was improved, after solving the cases with the improved MINLP model, better results were obtained, and raised the design feasibility. In order to overcome the complex pipeline connections problem between the equipment units which were brought by the timesharing design, a pipeline optimization model was proposed.The optimization results of the model showed that on the premise of given locations, it would not only minimize the total pipeline length, but also automatically gave the best location of the heat exchanger, optimized the network structure ultimately.