Search

In this work an innovative method for heat exchanger network (HEN) synthesis is introduced and examined. It combines a genetic algorithm (GA) with a heuristic based optimization procedure. The novel algorithm removes appearing heat load loops from the HEN structures when profitable, throughout the evolution. Two examples were examined with the new HEN synthesis method and for both better results were obtained. Thus, a positive effect of heuristic based optimization methods on the HEN synthesis with GA could be located. [ABSTRACT FROM AUTHOR]

The special case of unity dispersive Mach number of the hyperbolic axial dispersion model is investigated as the more realistic and simpler alternative to the parabolic model with zero Mach number. Simple corrections to the mean temperature difference or to the heat transfer coefficients are derived as functions of the dispersive Peclet numbers. As an example the model is applied to a cascade of stirred tanks in overall counterflow arrangement. [ABSTRACT FROM AUTHOR]

promising method for the efficient design of large-scale heat exchanger networks is the genetic algorithm. The heat exchanger network is divided into sub-networks which are optimized by a hybrid genetic algorithm. In additional steps these sub-networks are optimized by a monogenetic algorithm. An example consisting of 39 process streams was considered in detail. Significant improvements were made in reduction of total annual costs and the number of heat exchangers. [ABSTRACT FROM AUTHOR]

Dealing with complex process units like dividing wall columns pushes the focus on the determination of suitable modelling approaches. For this purpose a nonequilibrium stage model is developed. The successful validation is achieved by an experimental investigation of fatty alcohol mixtures under vacuum condition at pilot scale. Aim is the recovery of high purity products. The proposed model predicts the product qualities and temperature profiles very well. [ABSTRACT FROM AUTHOR]