Ernest N. Bart and Joseph Kisutcza. Chemical Engineering, New Jersey Institute of Technology, University Heights, Newark, NJ 07102-1982
The author has been active in obtaining computer solutions for various separation processes, for example, computer solutions for binary distillation problems using McCabe-Thiele and Ponchon-Savarit methods for steady state operations. Batch distillation has also been examined. Some of these, however, can be obtained from other sources such as commercially licensed software such as Aspen„µ, Hysys„µ and Chemcad„µ. The cost of the license can be expensive and in certain special circumstances the solutions provided may not be totally satisfactory. Several programs for solving batch distillation problems have been generated using Mathcad„µ 2001. The purpose of generating these programs was partially for teaching purposes although design usage is possible. Two basic types of distillation were considered. In the first type, the distillate composition is maintained constant and the reflux ratio is varied with time, and in the second type, the reflux ratio is fixed and the distillate composition is allowed to vary with time. For each of these types, a program was generated that could treat 26 different binary systems (systems without azeotropes). Each system was assigned a number that was connected to information in a database. By selecting the proper number, all the necessary information for that system would load. Additional programs for treating systems with azeotropes (maximum and minimum boiling) are currently being developed. In all cases, the number of ideal stages for the column and the feed composition must be specified, although the programs are provided with default values. For all systems the operating line and the stages may be animated to illustrate the course of events as a function of time. The programs will provide numerical output and graphical output as a function of time. The activity coefficients are generated by means of the Wilson equation and the vapor pressures by means of the Antoine equation. These computations are very difficult to achieve by hand because it involves fitting stages to an operating line that isn't known. This requires graphical trial and error procedures that are very time consuming, particularly if numerous stages are involved. Graphical integration is often needed. Here, the computer does all of this difficult work. The animation of the program is a wonderful aid to the process of teaching batch distillation, but also shows some aspects of the procedure not often considered. For example, consider a batch distillation in which the reflux ratio is held constant. Assuming that the column has five ideal stages, those stages at the start of the process will be doing roughly the same amount of enrichment. However, as time progresses, the lower end of the operating line may begin to approach the equilibrium curve. Thus, the bottom stages may have very small steps, resulting in very little enrichment. For such a situation, it may be that a column with fewer stages can do the same separation more economically. The systems used for this study were picked in order to obtain a wide range of relative volatility. The system ammonia - water has a very large relative volatility requiring very few stages. By contrast, the system carbon tetrachloride and benzene with a very low relative volatility usually requires large numbers of stages to achieve enrichment. Some the systems are ideal, such as benzene and toluene obey Raoult's law, whereas a system such as acetone and water is very non-ideal. The number of programs is in no way limited. Others may be readily added to the list.