Distillation dynamic behavior

Whereas there is extensive Hterature on design methods for azeotropic and extractive distillation, much less has been pubUshed on operabiUty and control. It is, however, widely recognized that azeotropic distillation columns are difficult to operate and control because these columns exhibit complex dynamic behavior and parametric sensitivity (2—11). In contrast, extractive distillations do not exhibit such complex behavior and even highly optimized columns are no more difficult to control than ordinary distillation columns producing high purity products (12). 

In the distillation column example, the manipulated variables correspond to all the process parameters that affect its dynamic behavior and they are normally set by the operator, for example, reflux ratio, column pressure, feed rate, etc. These variables could be constant or time varying. In both cases however, it is assumed that their values are known precisely. 

Many chemical and biological processes are multistage. Multistage processes include absorption towers, distillation columns, and batteries of continuous stirred tank reactors (CSTRs). These processes may be either cocurrent or countercurrent. The steady state of a multistage process is usually described by a set of linear equations that can be treated via matrices. On the other hand, the unsteady-state dynamic behavior of a multistage process is usually described by a set of ordinary differential equations that gives rise to a matrix differential equation. 

The dynamic behavior of processes (pipe-vessel combinations, heat exchangers, transport pipelines, furnaces, boilers, pumps, compressors, turbines, and distillation columns) can be described using simplified models composed of process gains, dead times, and process dynamics. 

In this section the methods developed in the previous section will be applied to analyze the dynamic behavior of integrated reaction separation processes. Emphasis is placed on reactive distillation and reactive chromatography. Finally, possible applications to other integrated reaction separation processes including membrane reactors and sorption-enhanced reaction processes will be briefly discussed. More details about reactive distillation processes were provided in Ref. [39]. For chromatographic reactors the reader should refer to Chapter 6 of this book, for sorption-enhanced reaction processes to Chapter 7, and for membrane reactors to Chapter 12. 

The developments in this subsection have revealed that high-purity distillation columns exhibit a dynamic behavior with three time scales. Thus, according to the results in Sections 7.4 and 3.4, the design of a control system involves the synthesis of a tiered structure featuring three levels of control action. 

It is worth mentioning that a set of the types of the dynamic behavior of the system in the case of copolymerization of m monomers is principally wider in comparison with the distillation process of an m-component liquid mixture as it has already been remarked [13]. The reason for this lies in the fact that copolymerization is a non-equilibrium process in contrast to distillation. In a particular case of three-component copolymerization such a possibility is shown... 

Even at steady state, efficiencies vary from component to component and with position in a column. Thus, if the column is not at steady state, then efficiencies also must vary with time as a result of changes to flow rates and composition inside the column. Thus, equilibrium-stage models with efficiencies should not be used to model the dynamic behavior of distillation and absorption columns. Nonequilibrium models for studying column dynamics are described hy, e.g., Kooijman and Taylor [AlChE 41, 1852 (1995)], Baur et al. [Chem. 

Let us now develop the state equations that will describe the dynamic behavior of a distillation column. The fundamental quantities are total mass and mass of component A. But the question is What is the system around which we will make the balances From a practical point of view,... 

All the equations above are state equations and describe the dynamic behavior of the distillation column. The state variables of the model are ... 

RD columns share some common features with chemical reactors on the one hand and with distillation columns on the other hand. The behavior of these multifunctional processes may be either close to that of non-RD columns or to chemical reactors. Further, new patterns of behavior can be introduced by the superposition of reaction and separation in a single processing unit. Hence, another interesting question that will be addressed in this chapter, is under what conditions and in what sense is the dynamic behavior of an RD column similar to that of a chemical reactor or to that of a non-RD column. 

In the previous case study, the focus was on control structure selection. As control algorithms standard linear Pl-controllers were used. In a second case study, the focus is on control algorithms. For that purpose we compare different control algorithms for a fixed control stmcture. The process to be considered is an industrial benchmark problem, which was treated in joint research with Bayer AG [21, 33]. The process and its open loop dynamic behavior is illustrated in Fig. 10.29. Components B and C are the reactants. They react in two consecutive equilibrium reactions to products A and E. The main product E is obtained in the bottoms of the column and the other product A in the distillate. 

Initially, most engineers wrote their own programs to solve both the nonlinear algebraic equations that describe the steady-state operation of a distillation column and to numerically integrate the nonlinear ordinary differential equations that describe its dynamic behavior. Many chemical and petroleum companies developed their own in-house steady-state process-simulation programs in which distillation was an important unit operation. Commercial steady-state simulators took over about two decades ago and now dominate the field. 

An important feature of all mass transfer operations and of a significant number of reaction systems in chemical engineering is the critical role played by interfacial phenomena. Liquid-liquid and gas-liquid systems are characterized by convective-diffusive transfer at interfaces that keep distorting (e g., in distillation, gas absorption, and liquid-liquid extraction). One fluid phase in such systems is often dispersed in another. The dynamic behavior of drops and bubbles, for example their shapes under various flow conditions and their breakage and coalescence, has been smdied for many years, one goal being to predict mass transfer rates in dispersed systems (Azbel, 1981 Clift et al., 1978 Mobius anad Miller, 1998). 

Answer. Improved residue curve mapping technique, multilevel modeling approach, dynamic optimization of spatial and control structures, steady-state and dynamic behavior analysis, generic lumped reactive distillation volume element, multiobjective optimization criteria. 

Damkohler number, 24 dilference point, 58 distillation line, 18 dynamic behavior, 112... 

Table 2 shows the lAE values obtained for each composition control loop of the distillation sequences under analysis. It is observed that the Petlyuk column offers the best dynamic behavior, which is reflected in the lowest values of lAE, for the control of the three product streams. The dynamic response of each control loop when the Petlyuk column was considered is displayed in Figure 2, where a comparison can be made to the response obtained with the widely-used direct sequence. One may notice in particular how the direct sequence is unable to control the composition of the intermediate component, while the Petlyuk column provides a smooth response, with a relatively short settling time. It is interesting to notice that for this mixture with an ESI = 1, and a low content of the intermediate component in the feed, the Petlyuk column offers the highest energy savings and also shows the best dynamic performance from the five distillation sequences under consideration. 

Similar equipment in series (for example extraction units in series) or chains of similar sections (for example trays in a distillation columns) or equipment in which variables are a function of time and location can be described dynamically by a section model in order to characterize the distributed character of the equipment. Typical dynamic behavior of a distributed system is ... 

Fig. 10.21. Liquid flow through a distillation column. The equations for the dynamic behavior become therefore ...

The dynamic response of the pressure to variations in C and H can be analyzed with the help of Fig. 34.3, which shows part of the interactions in the distillation colurtm. The response of the top vapor flow (Vj) on the cooling of the condenser (Q is usually relatively fast the time constant is determined by the heat capacity of the pipes (J/K), divided by the sum of the heat transfer coefficients and corresponding areas (W/K) inside and ontside the pipes. The evaporator (vapor flow Vh) will usually also show fast dynamic behavior. The difference between the vapor flow from the reboiler and to the condenser, yields after integration the response of the pressnre (the pressnre difference across the colnmn is ignored for reasons of simplicity). The speed of pressnre changes is limited by the total heat capacity of the trays, bottom and top. 

---