All the distillation columns considered up to this point in the book have used total condensers, where the distillate product is a liquid. However, many industrial columns have partial condensers in which the distillate product is removed as a vapor stream. This is commonly employed when there are very light components in the feed to the column that would require a high column pressure or a low condenser temperature to completely condense these very volatile components. The use of a partial condenser can avoid the use of costly refrigeration in the condenser.
The control of partial condenser columns is more complex because of the interaction among the pressure, reflux drum level, and tray temperature control loops. Both pressure and level in the reflux drum need to be controlled, and there are several manipulated variables available. The obvious are reflux flow, distillate flow, and condenser heat removal, but even reboiler heat input can be used. In this section, we explore three alternative control structures for this type of system, under two different design conditions: (1) a high vapor distillate flowrate (moderate reflux ratio) and (2) a very low vapor distillate flowrate (high reflux ratio).
As the dynamic simulation results will show, the preferred control structure depends on the control objectives of the entire process. For example, when the distillate goes to a downstream unit and wide variability in its flowrate is undesirable, the control structure should control pressure with condenser heat removal, control level with reflux, and maintain a constant reflux ratio.
The material in the section is taken from a paper1 in which English engineering units are used, so these are retained.
The numerical example used to study partial condensers is a depropanizer with a feed that contains a small amount of ethane but mostly propane, isobutane, and «-butane. Two cases are considered. The first has a feed composition that is 2 mol% ethane and 40 mol% propane, so the distillate flowrate is large and the reflux ratio is moderate (RR = 2.6). In the second case, the propane in the feed is only 4 mol% (with 0.02 mol% ethane), which gives a small vapor distillate flowrate and a large reflux ratio (RR = 20). Table 8.1 gives design parameters for the two cases. The Chao-Seader physical properties are used.
Design specifications are 1 mol% isobutane impurity in the distillate and 0.5 mol% propane impurity in the bottoms. The column contains 30 trays (32 stages) and is fed in the middle.
If the column is designed with a vapor distillate product, the column operates with a reflux drum pressure of 210 psia, which gives a reflux drum temperature of 110°F and permits the use of cooling water in the condenser. If a total condenser were used, the column pressure would have to be 230 psia to give a reflux drum temperature of 110°F. Of course, higher ethane concentrations in the feed would increase the difference between the operating pressures of total and partial condenser columns.
The major difference between the two cases is the distillate flowrate: 42.15 lb.mol/h in the first case and only 3.76 lb.mol/h in the second. The small vapor flowrate in the latter case corresponds to a volumetric flowrate of only 1.44 ft3/min. Considering the total volume of the 30-tray column (41.8 ft3) and the volume of vapor in the half-full reflux drum (9.5 ft3), the overall pressure time constant of the process is 36 min. This indicates that control of pressure using the small vapor distillate flow will be difficult and slow. The dynamic results given in a later section confirm this expected performance.
The diameters of the columns for the two cases are 1.6 and 1.5 ft, respectively. The reflux drum and base dimensions were sized to give 10-min holdups.
Was this article helpful?