Multi-stream Heat Exchanger Network Synthesis Based On The Pseudo-temperature

Heat transfers appropriately between hot and cold streams are extremely important which can increase the operating performance of the process. The chemical engineers are expected to analyse the topogoly of heat exchangers, the matches between hot and cold streams, the effect of heat exchangers' parameters on annual cost and offer the optimal approach. The pinch point is located by the stream heat transfer temperature difference contribution value instead of heat recovery approach temperature. Using parallelized GA/SA to optimize the temperature difference contribution value for every stream to obtain the near-optimum solution.Three factors having influence on the stream heat transfer temperature difference contribution value are analyzed deeply, then the equation of the stream heat transfer temperature difference contribution value is determined by statistical method. The stream pseudo-temperature incarnates the differences of streams in heat transfer. The stream apparent temperature is shifted into pseudo-temperature through the stream heat transfer temperature difference contribution value. Pseudo-temperature means stream heat transfer ability level which considers differences between streams in heat transfer coefficient and materials of construction etc.The pinch point is located and utilities are determined based on the pseudo-temperature problem table algorithm, the hot and cold composite curves are divided into several blocks by cutting vertically at every point where the composite curves bend (kink). "Vertical matches" against the composite curves are performed as much as possible to achieve the minimum total area. Systems with "Threshold temperature" or multiple pinch points also are synthesized in this paper. Number of branches and maximum area in each multi-stream heat exchanger are employed as new restriction to meet the industrial practice.Evaluate the feasible interval for the stream heat transfer temperature difference contribution value. A NLP model whose objective function is the annual cost of the network is presented to optimize the stream heat transfer temperature difference contribution value for each stream, there are no binary variables or superstructure in this model, furthermore, the number of continuous variables is smaller than the previous method. PGA/S A is used to solve the NLP problem. Design OCX operator for preserving paternal good information and EC operator for holding the diversities of population individuals, moreover, the idea of parallelalgorithm is introduced to advance the search capability and calculation efficiency of genetic algorithm. Finally, PGA/SA and matches in T-H graph based on the pseudo-temperature are successfully combined to synthesize the heat exchanger network. A group of two stream and multi-stream heat exchangers in series are obtained.