| Simulation is the key to analyze complex industry processes intensively. There are many un-stable processes in the industry systems, the essence of which can be studied only by dynamic simulations. The equation oriented dynamic modeling introduces the time variable into the model system, i.e., the characteristics are variating with time. The modeling of a dynamic process is a combination of process simulation, control theory, chemical unit models and thermodynamics. It is a system of differential algebraic equations(DAEs).This dissertation focuses on the dynamic simulations for different processes. The commercial software gPROMS is used as the modeling tool and simulation environment, and the algorithms are called for calculating the large-scale DAEs. The results of the dynamic and steady state simulations for several different kinds of processes are compared and analyzed.The main contributions include the following aspects:· The key problems in dynamic simulations with gPROMS are studied. Several examples are presented and discussed to answer the following 3 questions: how to call foreign object for the calculation of physical properties? What are the differences between different algorithms for different equation models? And what are the key factors that influence the dynamic simulation processes?· Equation oriented dynamic model is constructed for the multi-column system of depropanizer and debutanizer in ethylene separation plant. The dynamic simulations are focused on the start-up process and the switching processes between different working points. And the results of dynamic simulations are compared with the results of steady simulations using Aspen Plus.· Four kinds of different propositions about the relationship between the steady and the dynamic states are presented. The relationship is studied using differentkinds of processes as examples. Different models are established according to the four kinds of propositions, and simulation results are compared and analyzed. |
PDF Full Text Request