Optimum sequencing of columns

The literature of optimum sequencing of columns is referenced by King (1980, pp. 711-720) and Henley and Seader (1981, pp. 527-555). For preliminary selection of near optimal sequences, several rules can be stated as guides, although some conflicts may arise between recommendations based on the individual rules. Any recommended cases then may need economic evaluations. 

Porter, K. E., and Momoh, S. O., Finding the Optimum Sequence of Distillation Columns—An Equation to Replace the Rules of Thumb (Heuristics), Chem. Engg. J., 46 97, 1991. 

Heaven, D.L., Optimum Sequencing of Distillation Columns in Multicomponent Fractionation, M.S. Thesis, University of California, Berkeley, CA., 1970. 

The economics of the various methods that are employed to sequence multicomponent columns have been studied. For example, the separation of three-, four-, and five-component mixtures has been considered (44) where the heuristics (rules of thumb) developed by earlier investigators were examined and an economic analysis of various methods of sequencing the columns was made. The study of sequencing of multicomponent columns is part of a broader field, process synthesis, which attempts to formalize and develop strategies for the optimum overall process (45) (see Separation systems synthesis). 

The use of a column coupling configuration of the separation unit provides the possibility of applying a sequence of two leading electrolytes in one analytical run. Therefore, the choice of optimum separation conditions can be advantageously divided into two steps ... 

Application. Previous methods (Secs. 3.2.1 to 3.2.6) produce a design. They take product compositions and deliver the number of stages, reflux, and optimum feed stage. The Smith-Brinkley method rates a column using the reverse sequence of steps. It takes the number of stages, reflux ratio, and actual feed location, and yields the product compositions. 

To improve the stage distribution for each section in the crude tower, the tower is decomposed into a sequence of simple columns as a first step. After that, the ideal number of stages and the optimum feed stage for each simple column are found. Finally, the simple columns are merged back to the complex column with new numbers of stages for each section. 

In a three column system, on Column One the column operational sequence will be sample, flush, equilibrate, and then on Column Two one will have to equilibrate, sample, flush and to set up Column Three in sequence, one is going to equilibrate through the first two steps and then load sample and then flush. Note that the product is continuously coming out of the system after each sequence. This is different than if one had a single column and one had to wait for longer periods of time between events. It allows optimum utilization of downstream recovery equipment with less batch operation. 

The synthesis of optimum sequences for the multicomponent azeotropic mixture is the issue of the distillation theory. Geometric theory of distillation overcomes the principal part of this problem - the determination of possible splits for each potential distillation column that may be included into the synthesized sequence. The best feasible sequences selection is carried out on the basis of the criteria of a minimum number of columns, as well as minimum liquid and vapor flows, under the minimum reflux mode. 

The analysis of the minimum reflux mode is used at the stage of sequence selection, as well as at the stage of determination of optimum reflux ratios and the quantity of column trays. The geometric theory of distillation makes it possible to develop the general methods of calculation of minimum and more reflux mode. 

After identification of several preferable sequences, choosing among the optimum sequences, taking into consideration possible thermodinamic improvements and thermal integration of columns, arises. This task is similar to the synthesis of separation flowsheets of zeotropic mixtures (see Section 8.3), and it should be solved by the same methods (i.e., by means of comparative estimation of expenditures on separation). The methods of design calculation, described in Chapters 5 7 for the modes of minimum reflux and reflux bigger than minimum, have to be used for this purpose. In contrast to zeotropic mixtures, the set of alternative preferable sequences for azeotropic mixtures that sharply decreases the volume of necessary calculation is much smaller. 

The optimum process for this binary separation would be to have fixed positions for the introduction of mobile phase and feed, and fixed collection points for the two components of the mixture whilst having the ability to move the stationary phase upwards. In practice it is impossible to engineer a system where the column bed moves, but it is possible to simulate the movement. Such a system is shown schematically in Figure 1.6 where four columns are set in sequence with four multi-port valves between the columns. 

The column sequence could be reversed and other configurations could be laid out for different feed compositions. For instance, if Xp < Azi, then the feed must go to the high-pressure column first so that the resulting azeotropic composition will be on the other side of the lower-pressure azeotropic composition. If Xp > A z2> the feed must first be sent to the lower-pressure column. The final selection of a particular configuration among possible alternatives is best made on the basis of computer simulation of the various options to determine the economic optimum. 

Scale-up of chromatographic separations usually is done by testing at several sizes following a sequence such as analytical column. I in. (245 mm) diameter preparative column, 6 in. (15 cm) diameter eolumn. 3 ft (1 m) diameter column, with optimum injection-eJution policy determined at each step. Each scale of operation has its own distribation and packing method problems associated with it, which can ba solved only partially by resting at a smaller scale. 

---