HETP and pressure drop prediction for structured packing distillation columns using a neural network model

A neural net framework was used to predict the mass-transfer and hydraulic performance of a commercial structured packing operating in distillation se rvice. The results indicated that the approach produced a more accurate pre diction than a traditional semiempirical model. The neural net methodology was also used to yield a detailed sensitivity analysis of the operating var iables.