Modeling and Optimization of a Crude Distillation Unit: a case Study for Undergraduate Students
Download 244.37 Kb. Pdf ko'rish
|
Modeling and optimization of a crude dis
- Bu sahifa navigatsiya:
- Optimization
- Sensitivity Analysis.
- Optimization Process.
Results. A completely specified system has only one set of values
that solve all model equations and product specifications simul- taneously. In order to demonstrate that the system has only one valid solution, different sets of data can be used for initialization and they should lead to the same results. This is a way to confirm what is found in the degrees of freedom analysis; the model is completely specified by having the same number of variables and independent equations. The fact that a solution is found implies that all the specifications are satisfied at the same time; this is not always achievable considering the relationship between the operation conditions and the quality specifications. The temperature, liquid and vapor flow profiles, should keep the same tendency seen in the previous simulation. Optimization Before the process of optimization, it is advisable to have an idea of the effect and magnitude of the changes produced by the vari- ables in the objective function. For this, it is suggested to make a sensitivity analysis of each variable. Sensitivity Analysis. In order to evaluate the decision variables, each one are changed individually in a rage equivalent to 85% and 115% of its original value in the completely specified system. The answer to the analysis is the value of the objective function (flow of Stab-LSR in lb mol/s). The results presented in Table 8 show the variables that significantly affect the values of the objec- tive function: Steam flow, Steam-1 flow, Steam-2 flow, Steam-3 flow, Wet-Gas flow, reflux rate of the atmospheric distillation col- umn, and the condenser temperature of the stabilization column. The others the variables might be discarded for the optimization process. Optimization Process. The optimization is done using the tool provided with the software for this purpose (Optimizer). The ranges of variation of the chosen variables should be established. The ranges used in this optimization shown in Table 9. The optimization can be carried out analyzing the optimum value for each individual variable; subsequently, the variables can be merged in groups of two, three, and so on until they are all grouped. This process allows demonstrating the effect that adding variables and therefore enlarging the search region has on the objective value optimum. Results. Out of the seven variables chosen as manipulated, six of them (Steam flow, Steam-1 flow, Steam-2 flow, Steam-3 flow, Wet- |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling