STUDIES ON COMPUTER-AIDED MODELING, SIMULATION AND OPTIMIZATION IN-PROCESS DEVELOPMENT

This thesis is based on eight publications on computer aided modelling , simulation and optimization in process development. Modelling, simul ation and optimization of whole process plants are discussed. The meth ods by which opponent requirements in the plant are taken into account are reviewed. Similarly the subject of process synthesis is covered a nd its importance in process development emphasized. The methods by wh ich process synthesis can be included in optimization problems are dis cussed. In the process development phase it is common to create many a lternative process schemes. Thus there is a need for fast evaluation o f the investment costs of these alternatives. It is desirable to use t he same tool, a flowsheeting program, to create the process material a nd energy balances and to carry out the economical evaluation. A compa rative study with computer code implementations of investment cost est imation methods was done. Normally such things as process safety, cont rollability and economics can not be expressed in exact numbers. There fore to model and to simulate such systems the qualitative variables m ust be quantisized. One promising method is to use fuzzy sets together with linguistic variables. This thesis includes a review article on t he use of fuzzy sets and linguistic variables to handle data and knowl edge of uncertain and/or qualitative nature. The computer program IFTH EN for modelling of such systems is described. An application demonstr ates this methodology and the use of the computer program IFTHEN along with other software on the control and safety system of high pressure polyethene production. Often one of the important issues in the use o f modem computer programs developed for plant design is the selection of the most suitable thermodynamic models for the problem in hand. Thi s area is discussed on the basis of the classes of the needed properti es. The modelling of slurry polymerization of propene is discussed. A model for the polymerization kinetics was developed. This kinetic mode l was included in a slurry polymerization model of a continuous stirre d tank reactor system. This was transferred to a module into a general purpose flowsheeting program. The flowsheeting program has optimizati on capabilities which were used in this study to find out the most pro mising starting values for test runs at a full scale plant. The advant ages of this kind of modelling, simulation and optimization are review ed.