Distillation columns behavioral model

The modeling of RD processes is illustrated with the heterogenously catalyzed synthesis of methyl acetate and MTBE. The complex character of reactive distillation processes requires a detailed mathematical description of the interaction of mass transfer and chemical reaction and the dynamic column behavior. The most detailed model is based on a rigorous dynamic rate-based approach that takes into account diffusional interactions via the Maxwell-Stefan equations and overall reaction kinetics for the determination of the total conversion. All major influences of the column internals and the periphery can be considered by this approach. 

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 what follows, we begin by introducing two examples of process systems with recycle and purge. First, we analyze the case of a reactor with gas effluent connected via a gas recycle stream to a condenser, and a purge stream used to remove the light impurity present in the feed. In the second case, the products of a liquid-phase reactor are separated by a distillation column. The bottoms of the column are recycled to the reactor, and the trace heavy impurity present in the feed stream is removed via a liquid purge stream. We show that, in both cases, the dynamics of the system is modeled by a system of stiff ODEs that can, potentially, exhibit a two-time-scale behavior. 

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. 

Relaxation methods are not competitive with the steady-state methods in the use of computer time because of slow convergence. However, because they model the actual operation of the column, convergence should be achieved for all practical problems. The method has the potential of development for the study of the transient behavior of column designs and for the analysis and design of batch distillation columns. 

The prediction of the performance of three-phase multistage separation processes is dependent on the ability to describe the thermodynamics of three-phase behavior. The mathematical solution of three-phase distillation columns is similar to two-phase vapor-liquid columns, the difference being in the model used to calculate the /(-values. If the A -valuc model predicts two liquid phases, two liquid profiles must be considered in the column instead of one. 

The matrix Af of the component-material balances are of the same general form as the A, given by Eq. (2-18) for distillation columns. When plate 1 is assumed to behave according to model 1 (see Fig. 1-13) and plate /is assumed to have the behavior characterized by model 2 (see Fig. 2-2), the component-material balances may be represented by Eq. (2-18) provided that the elements of / are taken to be the following set... 

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

Do the same as in Problem III. 1 for the equations describing the dynamic and steady-state behavior of the binary distillation column modeled in Example 4.13. 

It has been shown (Popken et al, 2001 BeRling et al, 1998) that the first model usually provides the highest deviations between calculation results and the behavior of real reactive distillation columns. The assumption that chemical equilibrium is reached is not adequate for most of the chemical reactions of commercial interest. Changes in composition and heat are taken into account by using the equilibrium constant K and the heat of reaction. The second model caimot be recommended because chemical reactions are slower than the time needed to reach VLE. Therefore, it makes no sense to assume kinetic limitations for the distillation part but to neglect the reaction kinetics. 

Different physical modes are sometimes available for the same unit operation. A distillation column can, for example, be modeled on the basis of theoretical stages or using a rate-based model, taking into account the mass transfer on the column internals. A simulation of this kind can be used to extract the data for the design of the process equipment or to optimize the process itself During recent years, dynamic simulation has become more and more important. In this context, dynamic means that the particular input data can be varied with time so that the time-dependent behavior of the plant can be modeled and the efficiency of the process control can be evaluated. 

The impact of inaccurate -model parameters can be very serious. The parameters have a major influence on the investment and operating costs (number of stages, reflux ratio), The influence of the -model parameters on the results is especially large if the separation factor is close to unity. Poor parameters can either lead to the calculation of nonexisting azeotrojjes in zeotropic systems (see Section 11.1) or the calculation of zeotropic behavior in azeotropic systems. Poor parameters can also lead to a miscibility gap which does not exist." In the case of positive deviation from Raoult s law a separation problem often occurs at the top of the column, where the high boiler has to be removed, since at the top of a distillation column the most unfortunate separation factors are obtained. 

The behavior of process steps and phase systems can be represented by mathematical models (see the modeled-by relations in Fig. 1). In order to map the process topology to the model, process ports and connectors have to be related as well as process states and couplings by corresponding modeled-by relations. It must be considered that the relations between process design and model concepts are rather complex. Depending on the design context, a process step (such as separation) can be represented by different mathematical models (e.g. a linear splitter or a rigorous distillation column) and a certain mathematical model (such as a first-order transfer block) can be used to... 

A central issue is when is the model complete When correcting variables have not been taken into account, the system is usually undetermined. If important state variables have been forgotten, they will also not appear in the behavioral model and in any control scheme. For example, if, for simplicity reasons, in the case of distillation the column pressure has not been taken into account as state variable, the pressure will not appear in the behavioral model and pressure control will not be part of the system. 

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 ... 

Prominent examples include the exponential dependence of reaction rate on temperature (considered in Chapter 2), the nonlinear behavior of pH with flow rate of acid or base, and the asymmetric responses of distillate and bottoms compositions in a distillation column to changes in feed flow. Classical process control theory has been developed for linear processes, and its use, therefore, is restricted to linear approximations of the actual nonlinear processes. A linear approximation of a nonlinear steady-state model is most accurate near the point of linearization. The same is true for dynamic process models. Large changes in operating conditions for a nonlinear process cannot be approximated satisfactorily by linear expressions. 

The holdup effects can be neglected in a number of cases where this model approximates the column behavior accmately. This model provides a close approximation to the Rayleigh equation, and for complex systems (e.g., azeotropic systems) the synthesis procedures can be easily derived based on the simple distillation residue curve maps (trajectories of composition). However, note that this model involves an iterative solution of nonlinear plate-to-plate algebraic equations, which can be computationally less efficient than the rigorous model. 

The first step in the building the atmospheric distillation unit is entering the composition of the crude in order to generate the necessary hypothetical components for model. For the purposes of this simulation, we will consider the crude assays given in Table 2.5 to Table 2.8. It is important to remember that that we may have to remove extraneous details from the distillation curve to avoid unusual column behavior. We use the TB P distillation, density distribution and overall bulk density to define this system in Figure 2.14. 

---