Neural Network Models And Sensitivity Analysis For The Production Of Isopropyl Myristate In Semibatch Reactive Distillation
Isopropyl myristate (IPM) is an important chemical in the cosmetic and pharmaceutical industries. The IPM can be produced either through esterification or the transesterification process in semibatch reactive distillation (BRD). However, the latter process is not widely explored. The transesterifica...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2013
|
Subjects: | |
Online Access: | http://eprints.usm.my/43419/1/Nur%20Alwani%20Binti%20Ali%20Bashah24.pdf http://eprints.usm.my/43419/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Isopropyl myristate (IPM) is an important chemical in the cosmetic and pharmaceutical industries. The IPM can be produced either through esterification or the transesterification process in semibatch reactive distillation (BRD). However, the latter process is not widely explored. The transesterification process in BRD can be represented by a mathematical model, however, this model will end with a large number of differential equations and be very expensive to solve and will also be time consuming. Hence, the empirical model such as the artificial neural network (ANN) model provides better solution as it can deal with highly nonlinear and complex structures.
In this work, the production of industrial scaled IPM in BRD through the transesterification process is simulated using Aspen Plus and the simulation result achieved shows a comparable result as reported in the literature. The validated model is then used for sensitivity analysis to determine the relationship between the process input-output variables. The nonparametric test is used and the selected inputs are ranked according to their mean overall sensitivity. From the results, the reboiler duty, the initial mole of isopropanol, methyl mysistate, the reflux ratio, the feed flowrate and the temperature at stage 32 are considered as the input variables in the ANN model development to predict the bottom and distillate composition. |
---|