Before we get into the details of converting a steady-state simulation into a dynamic one, it might make sense to discuss some important steady-state calculations that are frequently performed to aid in the selection of a practical, effective control structure for a distillation column.
The majority of distillation columns are designed to attain a specified separation between the two key components. The two design degrees of freedom are usually specified to be the impurity of the heavy-key component in the distillate and the impurity of the light-key component in the bottoms. Therefore, in the operation and control of a distillation column, the "ideal" control structure would measure the compositions of the two products and manipulate two input variables (e.g., reflux flowrate and reboiler heat input) to maintain the desired amounts of the key component impurities in the two product streams.
However, very few distillation columns use this ideal control structure. There are a number of practical reasons for this. Composition analyzers are often expensive to purchase and have high maintenance costs. Their reliability is sometimes inadequate for online continuous control. They also introduce deadtime into the control loop if chromato-graphic methods are used. In addition, it is often possible to achieve very effective control without using direct composition measurements.
Temperatures are widely used to provide inferential control of compositions. Temperature sensors are inexpensive and reliable and introduce only small measurement lags in the control loop. In a binary system with constant pressure, temperature is uniquely related to composition. This is not true in multicomponent systems, but temperatures at appropriate locations in a distillation column can often provide fairly accurate information about the concentrations of the key components.
In "single end" control structures, the temperature on one tray is controlled. The remaining control degree of freedom is selected to provide the least amount of product
Distillation Design and Control Using Aspen™ Simulation, By William L. Luyben Copyright © 2006 John Wiley & Sons, Inc.
quality variability. For example, a constant reflux ratio RR can be maintained or the reflux to feed ratio R/F can be fixed. Sometimes the control of two temperatures is required (dual-temperature control). We discuss these alternatives in this chapter.
If tray temperatures are to be used, the issue is selecting the best tray or trays on which temperature is held constant. This problem has been discussed in the distillation literature for over a half-century, and several alternative methods have been proposed. We will review these alternative methods and illustrate their effectiveness for several systems.
It is important to note that all of these methods use only steady-state information, so steady-state process simulators such as Aspen Plus can be easily used to perform the calculations. The methods all require that various variables be held constant while other variables are changed. For example, two product compositions or a tray temperature and reflux flowrate may be held constant while feed composition is changed. The "Design Spec/ Vary" feature in Aspen Plus is used to fix the values of the desired independent variables and to calculate all the remaining dependent variables.
In several of the methods, the variable that is changed is the feed composition. The feed flowrate is not considered in any of the methods. This is because feed rate disturbances can be handled by ratioing the flowrates of the manipulated variable directly to the feed flow-rate. Maintaining a constant reflux ratio or a constant reflux to feed ratio should produce the same compositions and temperatures on all trays at any feed flowrate. Of course, this assumes that there are no changes in tray efficiencies with throughput. It also assumes that there are no changes in pressure on all the trays, which is seldom the case because of the changes in tray pressure drop and height of liquid on the trays as liquid and vapor rates change. But these effects are usually small enough that they do not adversely impact control to a great extent.
Was this article helpful?